Arachis ssp3 PDF

Feeding systems with forage legumes
to intensify dairy production in Latin

America and the Caribbean
A project executed by the Tropileche Consortium
editors
F. Holmann and C. Lascano
Feeding systems with forage

legumes to intensify dairy
production in Latin America
and the Caribbean
A project executed by the Tropileche Consortium
editors
2004
CGIAR System-wide Livestock Programme

P.O. Box 5689, Addis Ababa, Ethiopia
International Center for Tropical Agriculture

P.O. Box A.A. 6713, Cali, Colombia
International Livestock Research Institute

P.O. Box 30709, Nairobi, Kenya
I. Lascano, Carlos. II. Tropileche Consortium. III. International Livestock Research

Institute. IV. Title. V. Series.
Descriptors
1. Cattle 2. Multipurpose breeds 3. Small farms 4. Brachiaria decumbens 5. Arachis pintoi
6. Companion crops 7. Milk production 8. Proximate composition 9. Diet 10.
Supplements 11. Oryza sativa 12. Yields 13. Soil fertility 14. Cratylia argentea 15.
Hyparrhenia rufa 16. Silage 17. Zea mays 18. Glycine max 19. Brachiaria brizantha 20.
Stylosanthes guianensis 21. Innovation adoption 22. Marketing 23. Databases 24.
Diffusion of research 25. Latin America.
AGRIS subject category: L02 Animal feeding
LC classification: SF 203. H65
2004 ILRI (International Livestock Research Institute)

All rights reserved. Parts of this publication may be reproduced for non-commercial use
provided that such reproduction shall be subject to acknowledgment of ILRI as holder
of copyright.
ISBN 9291461041
Correct citation: Holmann F. and Lascano C. (eds). 2004. Feeding systems with
forage legumes to intensify dairy production in Latin America and the Caribbean: A project executed by the Tropileche Consortium. CIAT (Centro Internacional de Agricultura
Tropical), Tropileche Consortium, Cali, Colombia; SLP (System-wide Livestock
Programme), Addis Ababa, Ethiopia; and ILRI (International Livestock Research
Institute), Nairobi, Kenya. 172 pp.
Table of Contents
Project objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
List of abbreviations and acronyms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Executive summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Research results: Component 1Activity 1.1
Effects of Brachiaria decumbens alone and associated with Arachis pintoi
on milk production and milk components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
F. Romero and J. Gonzlez
Rice (Oriza sativa) yield as an indicator of soil fertility after the incorporation
of Stylosanthes guianensis stubble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
J. Vela
Evaluation of milk production systems using Cratylia argentea . . . . . . . . . . . . . . . . . . . . 17
C. Lascano, P. Toro and P. Avila
Nutritional complementarity between forage basal diets and legume supplements. . . . 20
W. Quionez, P. Avila and C. Lascano
Effects of dry season feeding of fresh and ensiled Cratylia argentea
on milk production and milk composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Substitution of chicken manure with Cratylia argentea in diets of cows grazing
Hyparrhenia rufa pastures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
M. Ibrahim, M. Franco, D. Pezo, A. Camero and J. Araya
Use of additives in Cratylia argentea silage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
C. Jimnez, L. Pineda and A. Medina
Planting maize in association with forage soybean for silage production:
Economic evaluation of partial sale of tender corn harvest . . . . . . . . . . . . . . . . . . . . 35
C. Jimnez, L. Pineda, B. Len and A. Montenegro
Planting maize in association with forage soybean for silage production:
Indicators of silage quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Milk urea nitrogen concentration for recommendations on optimum
protein-to-energy ratios in tropical forage diets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
P. Avila and C. Lascano
Tropileche Consortium
iii

Milk production of dual-purpose cows grazing Brachiaria brizantha
cv. La Libertad alone and associated with Arachis pintoi cv. El Porvenir . . . . . . . . . . 49
M. Lobo and V. Acua
Effect of regrowth age and cutting height on productivity of Cratylia argentea
cv. Veraniega . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
M. Lobo and V. Acua
Effect of supplementing fresh and ensiled Cratylia argentea cv. Veraniega
on milk production of dual-purpose cows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
M. Lobo and V. Acua
Effect of Centrosema macrocarpum on milk production . . . . . . . . . . . . . . . . . . . . . . . . . . 63
D. Lara and K. Retegui
Effect of associating Brachiaria brizantha with Arachis pintoi on milk yield . . . . . . . . . . . 67
D. Lara and K. Reategui
Use of Stylosanthes guianensis for strategic supplementation of pre-weaned calves . . . . . 71
J. Vela
Feeding pre-weaned calves with Stylosanthes guianensis in dual-purpose
production systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
J. Velasquez, G. Ruiz and C. Lascano
Decision tools to overcome nutritional constraints in dual-purpose cattle
in agro-silvopastoral systems: Net carbohydrate and protein system model . . . . . . . . 77
F. Holmann and R.W. Blake
Annual meeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
F. Holmann, C. Lascano and A. Ramrez
Annual workshop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
F. Holmann, C. Lascano, P. Kerridge and A. Ramirez
Regional consultation meeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
F. Holmann, P. Argel, C. Lascano and A. Ramrez
Ex ante analysis of new forage alternatives for dual-purpose cattle
farms in Peru, Costa Rica and Nicaragua . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
F. Holmann
iv
Feeding systems with forage legumes to intensify dairy production in LAC
Early adoption of Arachis pintoi in the humid tropics:

The case of dual-purpose cattle systems in Caqueta, Colombia. . . . . . . . . . . . . . . . 104
L. Rivas and F. Holmann
The need for forage technologies in the Alto Mayo Region
of the Peruvian Amazon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
S. Fujisaka, D. Lara, K. Reategui, J. Montenegro, R. Ventura,
M. Diaz, R. Diaz and D. White
Potential benefits of new forage germplasm in dual-purpose cattle systems
in the dry tropics of Costa Rica, Honduras and Nicaragua . . . . . . . . . . . . . . . . . . . 114
F. Holmann
Analysis of milk market for small-scale artisan cheese factories in watersheds
of Honduras and Nicaragua involved in livestock production . . . . . . . . . . . . . . . . 132
F. Holmann
Tropileche newsletter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
F. Holmann
Database on dual-purpose cattle production systems . . . . . . . . . . . . . . . . . . . . . . . . . . 144
A. Medina and F. Holmann
Tropileche in the Internet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Videotape release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
F. Holmann, C. Lascano, P. Argel and R. Goyenaga
Publication of brochures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Cratylia argentea: A shrub legume for livestock production in the tropics . . . . . . . . . . 148
P. Argel, J. Gonzalez and M. Lobo
Brachiaria brizantha CIAT 26110: cv. Toledo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
P. Argel, J. Gonzalez and M. Lobo
Publications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
List of project proposals developed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Tropileche collaborators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Project objectives
The projects main objective is to use improved legume-based pastures to increase milk
production in smallholder dual-purpose systems, thus increasing income of dairy producers and ensuring the sustainability of their natural resources.
vi
List of abbreviations and acronyms

ADF
CATIE
CIAT
CNCPS
CODESU
CORPOICA
CP
CTAR
DEPAAM
DICTA
DM
ECAG
FUNDAAM
GTZ
ha
IDB
IDR
IIAP
ILRI
IVDMD
IVITA
LAC
LW
MA
MAG
masl
ME
MUN
N
NDF
NFS
OM
R&D
SLP
TS
UCR
WFP
acid detergent fibre

Centro Agronmico Tropical de Investigacin y Educacin (Costa Rica)
International Center for Tropical Agriculture (Colombia)
Cornel Net Carbohydrate and Protein System
Consorcio para el Desarrollo de Ucayali (Peru)
Corporacin Colombiana de Investigacin Agropecuaria (Colombia)
crude protein
Consejo Transitorio Agropecuario Regional (Peru)
Desarrollo Participativo Amaznico (Peru)
Direccin de Ciencia y Tecnologa Agropecuaria (Honduras)
dry matter
Escuela Centroamericana de Ganadera (Costa Rica)
Fundacin para el Desarrollo Agropecuario del Alto Mayo (Peru)
German Agency for Technical Co-operation
hectare
Inter-American Development Bank
Instituto de Desarrollo Agrario (Nicaragua)
Instituto de Investigacin de la Amazona Peruana
International Livestock Research Institute
in vitro dry matter digestibility
Instituto Veterinario de Investigaciones Tropicales y de Altura (Peru)
Latin America and the Caribbean
animal liveweight
Ministry of Agriculture (Peru)
Ministry of Agriculture and Animal Husbandry (Costa Rica)
metre above sea level
metabolisable energy
milk urea nitrogen
nitrogen
neutral detergent fibre
non-fat solids
organic matter
Research and development
CGIAR System-wide Livestock Programme
total solids
University of Costa Rica
World Food Programme
vii
Executive summary
The project Feeding systems with forage legumes to intensify dairy production in Latin
America and the Caribbean is conducted under the CIAT-led Tropileche Consortium
that operates under the System-wide Livestock Programme (SLP) convened by ILRI
(International Livestock Research Institute). The Consortium is composed by scientists
from CIAT, ILRI, and national agricultural research institutes in Peru (FUNDAAM,
DEPAAM), Costa Rica (MAG, ECAG, CATIE, UCR), Nicaragua (IDR), and Honduras
(DICTA). The strategy to improve feeding systems includes:
evaluation of new feed resources to satisfy the nutritional requirements of grazing
animals
on-farm evaluation of new legume-based forage components and
economic analysis and acceptability/adoption studies of new technologies.
This report includes the results generated during the last four years of the project
(19962000), which further confirm the significant impact of improved grasses and
legumes on animal production.
In Costa Ricas Pacific coastal region, the pasture association Brachiaria decumbens and
Arachis pintoi increased the daily milk production of Jersey cows, even though these
animals were receiving commercial concentrates. Likewise, this association supported a
heavier stocking rate than did pastures of B. decumbens and Hyparrhenia rufa or of B.
Decumbens alone under similar management. In the same region, research on Cratylia
argentea (cratylia) demonstrated that this legume harvested fresh at 90 days after regrowth
and 90 cm height can replace commercial concentrate or 82% of chicken manure given
to dairy cows of medium production, and is an excellent feed alternative for the dry
season. The use of cratylia supplements could therefore help small milk producers have
access to a protein supplement produced on their own farms, thereby increasing their
income and cash flow.
Another form of using cratylia in the region is as silage with molasses added and fed
to cows, together with chopped sugarcane, during milking. Other studies found that
mixed cropping of maize with forage soybean is an excellent alternative for making silage,
not only because of the increase in protein in the final product, but also because of the
economic benefit from additional milk production and the partial use of harvested
maize through the sale of tender maize ears.
At the CIAT-Quilichao station (Department of Cauca, Colombia), researchers found
that about 10 mg/dL of urea in the milk comprised an adequate reference point at which
to increase the protein content in the animals diet. Dairy cows would most probably
respond to increased protein with significantly higher milk yields (provided that the
cows genetic potential is demonstrated).
In the Colombian Amazon, the use of Stylosanthes guianensis (stylo) is an attractive
option for raising pre-weaned calves and milking cows on small farms. The cost of establishing the legume is less than that of other alternatives based on grass and legume mix-
viii
Executive summary
tures. Moreover, cash flow increases because of increased sales of milk, without sacrificing calf weight gain. Because stylo can persist for 3 or 4 years, this technology can also
form part of a crop/pasture rotation system, eliminating the need to rest the land or
leave it in fallow. During rotation, the legume improves the soil through N fixation and
nutrient recycling.
In Pucallpa, Peru, stylo can fix as much as 50 kg/ha of N when it is established as a
pioneer crop for sowing rice and the later establishment of improved pastures. In the
Alto Mayo region of Peru, Centrosema macrocarpum has a positive impact when used as a
supplement feed for Holstein Gyr cows. Farmers in this region are highly likely to
adopt, with positive impact, the technologies developed by the Tropileche Consortium.
A high proportion of farmers had already adopted new cut-and-carry forages, improved
milking pens and sheds to provide feed supplements to the herd, and installed forage
choppers.
An ex ante study examined the impact on the herds of small dairy farmers of Costa
Rica, Honduras and Nicaragua of several forage alternatives, i.e.:
Level 1 = new germplasm based on cratylia and sugarcane;
Level 2 = same areas of cratylia and sugarcane as those of level 1, plus areas of the
Brachiaria varieties used in each country; and
Level 3 = areas of cratylia and sugarcane larger than those of levels 1 and 2, plus associations of Brachiaria + Arachis pastures.
The results were:
1. Level 1 permitted farmers in the different countries to completely eliminate the purchase of supplements for livestock for the dry season, thereby greatly improving the
farmers cash flow by reducing milk production costs by 11% in Nicaragua, 14% in
Costa Rica and 25% in Honduras.
2. Level 2 not only provided the benefits of level 1 but also helped maintain the same
milk production and herd size on smaller areas, releasing areas for alternative uses.
The released areas varied from 9% for Honduras, 39% for Costa Rica, to as much as
45% for Nicaragua. This level of adoption also reduced milk production costs compared with level 1 by 8% in Honduras, 11% in Costa Rica and 12% in Nicaragua.
3. Level 3 increased milk productivity/hectare to 964 kg in Nicaragua, 1390 kg in Costa
Rica and 1530 kg in Honduras.
The potential impact of the adoption of these technologies at the regional level is significant. For level 1, at the time of study, the resources invested totalled US$ 2.7 million
in Honduras, US$ 5.1 million in Nicaragua, and US$ 6.4 million in Costa Rica. By reducing production costs, this investment made possible an additional annual net earning
of US$ 9.9 million in Nicaragua, US$ 11.9 million in Costa Rica and US$ 12.6 million
in Honduras. Accordingly, the potential benefit of such investment is large, making
repayments possible within the year.
In Honduras and Nicaragua, the artisan-scale cheese industry is the principal buyer of
milk produced by small and medium-sized dairy farmers. In both countries, in watersheds with seasonal climates, milk production during the rainy season is almost double
that produced in the dry season, causing, respectively, an oversupply and scarcity of milk.
ix
A survey found that, during the dry season, even though the artisan-scale cheese producers were willing to buy 55% more milk in Honduras and 76% in Nicaragua, milk was
unavailable. The main reason was apparently the lack of adoption of improved foragebased feeding technologies.
This situation suggests that an aggressive program to promote the use of the shrub
legume C. argentea and sugarcane to supplement the herds feed during the dry season
would have much more impact than promoting grasses or legumes for the rainy season.
The artisan-scale cheese producers in both countries, but particularly in Honduras, require better quality milk, especially during the rainy season. In Honduras, milk prices
would be 9.4% higher during the dry season and 11.2% higher during the rainy season.
In Nicaragua, artisan-scale cheese producers would offer 17% higher prices, but only
during the rainy season.
In parallel with research, the Tropileche Consortium has developed media for communicating with its partners and users to disseminate research results. Media include a
database on dual-purpose production systems, which database is available through Internet
on the CIAT/Tropileche web site; the Hoja Informativa Tropileche (the Tropileche Newsletter),
which contains information on the Consortiums research advances; informative brochures on B. brizantha cv. Pasto Toledo; publication of results in scientific journals; and
videos on the impact of improved forage technologies.
In collaboration with Nicaraguas Instituto de Desarrollo Rural (IDR) and Honduras
Direccin de Ciencia y Tecnologa Agropecuaria (DICTA), the Consortium established
more than 65 ha of improved forages on 20 farms located at five sites in Nicaragua and
Honduras. Other achievements include the spontaneous adoption of C. argentea in Costa
Rica, where 230 kg of experimental seed and for commercial sowing were distributed to
seed companies and 41 farmers at four different sites; the commercial release in October
2000 in Costa Rica of B. brizantha CIAT 26110 as cultivar Pasto Toledo; and the expected
release of C. argentea CIAT 18516/18668 as cultivar Veraniega.
The challenge ahead is to facilitate widespread evaluation of herbaceous and shrub
legumes by small farmers and strengthen the production of seed of selected legumes. We
are identifying new research needs from problems that are being experienced by farmers
evaluating the new technologies. To effectively accomplish these objectives we need to
continue to strengthen the alliances with other ILRI and CIAT projects, with partners in
national R&D institutions, and the private livestock and seed sectors.
F. Holmann and C.E. Lascano
Achievements attained for each component and activity of

the Tropileche Consortium, 19962000
Component
Activity
Outstanding achievements
1.1 Controlled feeding and

grazing trials to
determine the
relationships between
milk production and
forage resources
The use of forage legumes as protein supplements during

the rainy season increases milk production in cows with
high genetic potential
The use of forage legumes in association with improved
grasses increases the stocking rate and reduces weeds,
while improving the quality of biomass on offer in terms
of crude protein and digestibility and favouring pasture
sustainability
Pastures of grasses and legumes in association increase
the content of total solids in milk, resulting in better
prices for the farmer
1.2 Use of forage resources

and feed supplements
during the dry season
The response of cows in terms of milk production to

feed supplements of fresh or ensiled forage legumes +
sugarcane is similar to their response to protein sources
in commercial concentrates
Direct grazing of shrub legumes increases milk production, compared with the cut-and-carry system. The
legumes provide protein to the forage of the base diet,
which is usually deficient in nitrogen
Highly fermentable carbohydrates can stimulate the
degradation of less fermentable fibre
The use of carbohydrate additives improves the quality
of ensiled shrub legumes
1.3 Functional relationships

between forage resources
of contrasting quality,
ruminal fermentation,
and blood parameters
Concentration of milk urea nitrogen (MUN) as an

effective indicator of the protein:energy relationship in
the diet
Concentration of MUN to increase the efficiency of
forage resources on the farm
Optimising
forage use
Developing new 2.1 On-farm evaluation of

different forage systems
feed alternatives
with dual-purpose cattle
for dual-purpose
cattle
Forage systems for strategic feed supplementation during

the dry season reduce the need to buy feed concentrate
and increase the income of dual-purpose farms
The use of legumes with an energy source during the dry
season influences milk production
2.2 Evaluating new feed

alternatives to permit
early weaning
Fattening pre-weaned calves on legumes during the dry

season makes the selling of more milk possible and
increases animal liveweight gain as compared with
fattening calves on pastures of grass alone
2.3 Integration of
information, using
simulation models of
ruminants and nutrition
Simulation models are useful for reducing research costs,

predicting nutritional limitations and facilitating
decision-making on the priorities of animal nutrition
research
2.4 Workshops and meetings Workshops with the participation of the Tropileche
Consortiums partners facilitate co-ordination, increase
to analyse and plan
research efficiency, reduce the risks associated with
research activities
duplicated efforts, and permit the prioritisation of
medium- and long-term activities
(cont...)
xi

(cont)
Component
Activity
Outstanding achievements
3.1 Diagnosis, economic

characterisation and
land use in reference
sites
Ex ante analyses of new improved technologies to

identify constraints to the adoption of new technologies
by local farmers
Ex post analyses of new technologies to determine the
economic impact of new adopted technologies. A high
demand exists for new alternatives based on improved
forages to intensify milk production in the Alto Mayo
region of the Peruvian Amazon
The establishment of shrub legumes with sugarcane is a
starting point toward achieving greater economic impact
on small dual-purpose farms located in the dry tropics of
America
In Central America, the demand for milk from artisanal
cheese producers is higher during the dry season than in
the rainy season
The artisanal cheese producers of Central America are
willing to pay better prices for milk of better hygienic
quality
3.2 Disseminating research

results
The dissemination of results is integral to research and

technology transfer, and is essential in the adoption of
new technologies
xii
Usefulness of
the new forage
systems
Foreword
The main objective of the Tropileche Consortium is to increase milk and beef production of dual-purpose livestock in smallholder farms by developing improved foragebased feeding systems.
The value of livestock production in Latin America and the Caribbean (LAC) represents nearly 13% of the world production and 47% of the production in developing
countries. Milk and meat provide 20% of the protein consumed by the regions population. However, LAC has a 12% deficit in milk production, which is currently compensated by imports.
The livestock population in LAC is about 330 million head and nearly 78% of this
population is in the hands of small farmers with dual-purpose cattle systems (i.e. producing milk and beef from the same system). The dual-purpose cattle system accounts for
42% of the milk produced in the region. In addition, there are 590 million hectares of
pastureland, most of which is in advanced stage of degradation.
Tropileche is a CIAT-led consortium that operates under the System-wide Livestock
Programme (SLP) convened by ILRI (International Livestock Research Institute). The
Consortium initially selected the dry hillsides of Costa Rica (Central America) and the
forest margins of the Peruvian Amazon as benchmark sites for its activities. In these
areas, characterised by their fragility and severe erosion, livestock production is an important form of land use and a key economic activity for the wellbeing of the smallholder. The Consortium later extended its activities to the hillsides of Nicaragua and
Honduras and the forest margins of Colombia.
In the forest margins, degraded pastures, which lead to low-quality feed, limit milk
production. In the seasonally dry hillsides, the major constraint to increasing milk production is the lack of forage.
The following strategies were designed to overcome the above-mentioned constraints:
(a) an evaluation of new feed resources that meet nutritional requirements of grazing
animals, (b) on-farm participatory evaluation of new legume-based forage components to
intensify livestock production and enhance sustainable land use, and (c) characterisation
of benchmark sites, economic analysis and acceptability/adoption studies on new technologies.
The Consortium takes advantage of the existing capability of the following entities:
CIAT in forage germplasm development and land use characterisation; ILRI in livestock
production systems; the University of Cornell in nutrient optimisation models for ruminants; and national agricultural research organisations active at the projects benchmark
sites in Peru (FUNDAAM, DEPAAM), Costa Rica (MAG, ECAG, CATIE, UCR),
Nicaragua (MA), and Honduras (DICTA).
The Inter-American Development Bank (IDB) and the German Agency for Technical
Co-operation (GTZ) jointly finance the Tropileche Consortium project through the
CGIAR System-wide Livestock Programme, co-ordinated by ILRI (International
Livestock Research Institute).
This report presents the activities carried out for each research component carried
out at the benchmark sites.
Research results
Component 1
Optimising forage use
Activity 1.1 Controlled feeding and grazing trials

to determine the relationships between milk production
and forage resources
Highlights
The use of forage legumes as protein supplements during the rainy season increases
the milk yield of cows showing high genetic potential.

The use of forage legumes in association with improved grasses allows the stocking
rate to be increased and weed incidence to be reduced, while improving the quality of
edible biomass in terms of crude protein and digestibility, thus making the system
more sustainable.
Pastures of grasses and legumes in association increase the total solids content of
milk, increasing the price obtained by the producer.
Effects of Brachiaria decumbens alone

and associated with Arachis pintoi on milk
production and milk components
ECAG, Costa Rica
Arachis pintoi was initially used in association with grasses in the humid tropics. One of
the strategies proposed by the Tropileche Consortium for Costa Rica was to search for
alternatives to improve livestock feeding in those regions where producers find it difficult to maintain a sustainable production during the dry season. One of these alternatives is the use of Bracharia decumbens/Arachis pintoi pastures.
Objectives
To evaluate milk production and milk components in cows grazing a pasture of
Brachiaria decumbens alone, B. decumbens associated with A. pintoi and B. decumbens

mixed with Hyparrhenia rufa.
To evaluate the availability, quality and dynamics of the botanical composition of the
above-mentioned pastures.
Materials and methods

The experiment was carried out at the Escuela Centroamericana de Ganadera (ECAG),
located in Balsa de Atenas, Costa Rica, at an altitude of 460 metres above sea level
(masl). The mean annual temperature is 28C, with an average annual precipitation of
1500 mm and well-defined dry season (December to May) and rainy season (June to
November).
A degraded pasture of B. decumbens invaded by jaragua grass (H. rufa), a pasture of B.
decumbens in good condition, and a 10-years-old pasture of B. decumbens associated with
A. pintoi (CIAT 18744) were used. Each pasture was divided into two paddocks (1.25 ha
each). One paddock was used for animal adjustment (7 days) and the other for measurement (7 days). Grazing consisted of 35 days of rest and 7 days of occupation.
The experiment began in October 1995 with the establishment of A. pintoi in a degraded B. decumbens pasture using plant material sown in furrows spaced 1 m apart. The
pasture was first grazed at 30 days, and evaluations began in July 1996 and finished in
July 2000. Every year, during the rainy season, samplings were carried out in the pad-
docks to measure botanical composition, forage availability and quality of forage on

offer. These measurements were performed on Day 1 of the measurement period.
Jersey cows of the ECAG herd, of different categories and stages of lactation, were
used. Their milk production was measured during the 7 days of the measurement
period, and milk components were measured on Day 4.
Results
Forage availability
Table 1 indicates that more forage was available in the associated pasture than in the
other pastures evaluated. Results indicate that 33% more forage was available in the
associated pasture as compared with the B. decumbens + H. rufa pasture and 39% more
than that of the pasture of B. decumbens alone.
Table 1. Forage availability in terms of dry matter (DM), crude protein (CP) content and
in vitro dry matter digestibility (IVDMD) of pastures evaluated in ECAG, Costa Rica.
Pastures
DM
(t/ha per cycle*)
CP
(%)
IVDMD
(%)
Brachiaria decumbens + Arachis pintoi
3.2a**
14.0a
69.3a
B. decumbens + Hyparrhenia rufa
2.4b
10.2b
63.2b
B. decumbens
2.3b
11.4b
67.8a
Significance (P < 0.05)
0.0018
0.0002
0.0001
* Grazing cycle every 3035 days.

** Averages with the same letters within the column do not differ significantly (P < 0.05),
according to Duncans test.
In addition to the increase in availability of dry matter (DM), the inclusion of the
legume had a significant effect (P < 0.05) on crude protein (CP) content and on in vitro
dry matter digestibility (IVDMD) of pastures on offer. In the first case, an increase of
approximately 3% was obtained and in the second, more than 4%. van Heurck (1990)
and Gonzlez (1992) obtained similar results, in pastures of African star grass (Cynodon
nlemfuensis) alone and associated with A. pintoi.
Stocking rate
In 1996, at the beginning of the trial, the same stocking rate was used in all pastures
(Table 2). In 1997, the stocking rate was increased in the B. decumbens + A. pintoi pasture
to 3.2 animal unit (AU/ha) but was maintained the same as in the other pastures (2.5
AU/ha). During 1998 and 1999 the excellent conditions of the associated legume pasture allowed stocking rates above 4 AU/ha, which was almost 1 AU/ha higher than that
of the other pastures.
It should be highlighted that these results correspond to the stocking rate used during the rainy season and that cows consumed commercial concentrate at a ratio of 3:1 (3
Effects of B. decumbens alone and associated with A. pintoi on milk production
kg concentrate per litre of milk produced). However, it is expected that this increase in
stocking rate in non-associated pastures could be non-sustainable over time because of its
accelerated degradation.
Table 2. Stocking rate (AU/ha)1 of pastures evaluated at ECAG, Costa Rica.
Evaluation periods (years)
Pastures
1996
1997
1998
1999
Average
2.5
3.2
4.4
4.6
4.6
2.5
2.5
3.7
3.6
3.6
B. decumbens
2.5
2.6
3.5
3.7
3.7
1. 1 AU = 450 kg animal liveweight.
Botanical composition
Figure 1 shows that the botanical composition of the associated pasture increased from
23% for A. pintoi in year 1 to 43% in subsequent years of grazing. The percentage of B.
decumbens was constant (59%), whereas the percentage of weeds decreased. Between 1996
and 1998, both grasses in the B. decumbens + H. rufa pasture increased slightly (5% for B.
decumbens and 8% for H. rufa). Weed incidence decreased from 19% to 6% during the
same period. However, in 1999, the percentage of both grasses tended to decrease while
that of weeds increased.
Incidence (%)
70
60
50
40
30
20
10
0
1
2
3
Years (between 1996 and 1999)
B. decumbens
A. pintoi
Weeds
Figure 1. Botanical composition of the pasture Brachiaria decumbens + Arachis pintoi (ECAG, Costa Rica).
Between 1996 and 1997, in the grass pasture alone, the percentage of B. decumbens increased from 70 to 94%, whereas those of other pasture components decreased (Figure 3).
In subsequent periods, the percentage of B. decumbens decreased from 94% to 83%, and
weeds increased by 10%.
Incidence (%)
70
60
50
40
30
20
10
0
1
2
3
B. decumbens
H. rufa
Malezas
Figure 2. Botanical composition of the pasture Brachiaria decumbens/Hyparrhenia rufa (ECAG, Costa Rica).
Incidence (%)
100
90
80
70
60
50
40
30
20
10
0
1
2
3
B. decumbens
Other grasses
Weeds
Figure 3. Botanical composition of a pasture of Brachiaria decumbens alone (ECAG, Costa Rica).
The botanical composition of the pastures partially explains the changes in stocking
rate during the experimental period (Table 2). Therefore, as of 1997, the components of
the associated pasture finally stabilised and allowed the stocking rate to be increased to
4.6 AU/ha. In 1998, the increased biomass of the B. decumbens + H. rufa pasture allowed
the stocking rate to be increased by 1 AU/ha, reaching 3.6 AU/ha. This stocking rate,
however, was apparently too high and favoured pasture degradation.
These results confirm that pastures of grasses and legumes in association can support
higher stocking rates than non-associated pastures, without affecting their botanical
composition and quality and without increasing the presence of weeds.
Milk production and composition

Table 3 presents milk production/cow per day in the different treatments. The highest
production average (12.4 kg/cow per day) was found in the pasture of B. decumbens
associated with A. pintoi, being significantly higher (P < 0.011) than those obtained in
pastures of B. decumbens + H. rufa (11.7 kg/cow per day) and B. decumbens alone (11.6 kg/
cow per day). These data agree with those obtained by van Heurck (1990) and Gonzlez
(1992), in which the highest milk production was found in a pasture of African star grass
associated with A. pintoi as compared with a pasture of African star grass alone.
Table 3. Effect of three types of pasture on milk production and milk components in Jersey cows receiving a commercial
supplement (ECAG, Costa Rica).
Pastures
Production
(kg/cow per
day)
Fat
(%)
Protein
(%)
Lactose
(%)
Total
solids
(%)
Non-fat
solids
(%)
9.07a
Brachiaria decumbens + Arachis

pintoi
12.4a*
4.84a
3.66a
4.68a
13.89a
11.7b
4.71b
3.58ab
4.70a
13.63c
8.96a
B. decumbens
11.6b
4.85a
3.54b
4.62b
13.73b
8.89a
0.0004
0.0562
0.0381
Significance (P < 0.05)
0.011
0.0004
0.1154
* Averages in the same column followed by the same letter do not differ significantly (P < 0.05), according to Duncans
test.
It is important to highlight that, although the Jersey cows used in the trial received as
supplement a feed concentrate of 14% CP and 2.4 Mcal ME, the beneficial effect of
Arachis on milk production could still be observed.
Table 3 shows the components of the milk obtained from the different pastures. A
higher fat content (P < 0.0004) was found in milk from cows grazing pastures of B.
decumbens associated with A. pintoi (4.84%) and B. decumbens alone (4.85%). This value
for the pasture of B. decumbens + H. rufa was 4.71%. Cipagauta et al. (1998) found
similar milk fat values in B. decumbens + A. pintoi pastures in the Colombian Amazon
region, but fat content of milk produced by cows grazing B. decumbens alone was even
lower than the native pasture. These results, however, differed with those found by
Gonzlez (1992) in which the percentage of fat did not differ among the three pastures
evaluated.
The milk protein content varied among pastures, and was higher in the pasture associated with A. pintoi (3.66%), compared with the pasture B. decumbens + H. rufa
(3.58%) and B. decumbens alone (3.54%).
Similar values of lactose were found in the milk of cows grazing the B. decumbens + A.
pintoi pasture (4.68%) and B. decumbens + H. rufa (4.70%), but were higher (P < 0.0381)
to that of milk of cows grazing the pasture of B. decumbens alone (4.62%).
The total solids content of milk varied among pastures, being highest (P < 0.0004) in
the pasture of B. decumbens associated with A. pintoi (13.89%), followed by B. decumbens
alone (13.73%) and B. decumbens associated with H. rufa (13.63%). The levels of non-fat
solids (NFS) in milk did not differ significantly among the three pastures.
The effect of the treatments (pastures), without the inclusion of a commercial feed
supplement, on milk production and milk components was evaluated over one year in
the same Jersey cows. Table 4 shows the significant difference (P < 0.0279) in milk production/cow per day, which was higher in the pasture associated with A. pintoi (9.3 kg/
cow per day) than in the other pastures. This difference is similar to that obtained with
cows fed concentrate at a ratio of 3:1. No differences were found among the three treatments regarding milk components; however, these components differed from those
found in the group of cows fed concentrate.
Table 4. Effect of three types of pasture on milk production and milk components of Jersey cows that did not receive a
dietary supplement (ECAG, Costa Rica).
Pasture
Brachiaria decumbens +
Arachis pintoi
Production
(kg/cow per day)
9.3a*
Fat
(%)
Protein
(%)
Lactose
(%)
5.4
3.7
4.6
4.7
3.7
4.6
B. decumbens +
Hyparrhenia rufa
8.1b
B. decumbens
8.3b
5.3
3.8
4.6
Significance
0.0279
0.5
0.5
0.5
Total solids
(%)
14.4
Non-fat
solids (%)
9.1
13.7
9.0
14.4
0.3971
9.1
0.5
* Averages in the same column followed by the same letter do not differ significantly (P < 0.05), according to the Duncans
test.
Table 5 shows the availability and quality of the grass offered to cows that were not
fed concentrate. The difference observed in milk production can be attributed to the
greater availability of forage and its better quality in the associated pasture, as was the
case of the pasture of B. decumbens associated with A. Pintoi.
Conclusions
Based on study results, we conclude that:
Pastures of B. decumbens associated with A. pintoi increase the daily milk production
of Jersey cows, even when they receive a commercial feed concentrate. These pastures
are also capable of supporting a higher stocking rate than B. decumbens + H. rufa pastures and B. decumbens alone, under similar management conditions.
10
Table 5. Availability and quality of forage offered to Jersey cows that were not fed a commercial concentrate
(ECAG, Costa Rica).
Pastures
DM
(t/ha per grazing cycle*)
CP
(%)
IVDMD
(%)
3.9a**
14.2a
63.3a
3.4b
10.1b
62.5a
B. decumbens
2.7c
11.7b
63.0a
Significance
0.0018
0.0066
0.9058
* Grazing cycles of 3035 days.

** Averages in the same column followed by the same letter do not differ significantly (P < 0.05), according to
Duncans test.
The persistence of A. pintoi during the four years of the evaluation contributed to
maintaining production and pasture quality, confirming the high potential of this
legume in hillside areas with defined dry season in Central America.
References
Cipagauta M., Velasquez J. and Pulido J.I. 1998. Produccin de leche en tres pasturas del
piedemonte amaznico del Caquet, Colombia. Pasturas Tropicales 20(3):29.
Gonzlez M.S. 1992. Selectividad y produccin de leche en pasturas de estrella africana (Cynodon
nlemfuensis) solo y asociado con las leguminosas Arachis pintoi CIAT 17434 y Desmodium
ovadifolium CIAT 350. MSc Thesis. CATIE (Centro Agronmico Tropical de Investigacin y
Educacin), Turrialba, Costa Rica.
Gonzlez M.S., van Heurck L.M., Romero F., Pezo D.A. and Argel P.J. 1996. Produccin de leche
en pasturas de estrella africana (Cynodon nlemfuensis) solo y asociado con Arachis pintoi y
Desmodium ovadifolium. Pasturas Tropicales 18(1):212.
van Heurck L.M. 1990. Evaluacin del pasto estrella (Cynodon nlemfuensis) solo y asociado con las
leguminosas forrajeras Arachis pintoi CIAT 17434 y Desmodium ovalifolium CIAT 350 en la
produccin de leche y sus componentes. MSc Thesis. CATIE (Centro Agronmico Tropical
de Investigacin y Educacin), Turrialba, Costa Rica.
11
Rice (Oriza sativa) yield as an indicator

of soil fertility after the incorporation
of Stylosanthes guianensis stubble
J. Vela
IIAP, Peru
Stylosanthes guianensis (stylo) is a legume adapted to the humid rainforest conditions of

the Peruvian Amazon region, where it presents good growth and high biomass production. Smallholders in this region commonly plant rice (Oriza sativa). However, for
this crop to develop well, it requires the application of N, which is a high-cost input.
This study aimed to evaluate the potential of stylo to fix nitrogen and determine the use
of this element by the rice crop prior to pasture establishment.

The experiment was carried out between October 1998 and February 1999 on the Sara
farm, located 15 km off the road that leads from Pucallpa to Tingo Mara, on an acid
Ultisol with high aluminium content. This area belongs to the seasonal semi-evergreen
tropical rainforest ecosystem with annual mean precipitation of 1900 mm and mean
temperature of 26C.
Treatments were as follows:
T1, soil without application of N (control)
T2, soil after incorporating S. guianensis stubble without application of N
T3, T4, T5 and T6, soil after incorporating S. guianensis stubble with the application
of 50, 100, 150 and 200 kg/ha of N, respectively.
The variables evaluated were rice yield; number of tillers and of spikelets/m2, number
of grains/spikelet and weight of 1000 grains. Rice variety Chancabanco, with a 90-day
vegetative period, was planted. All treatments received a uniform basal application of 50
kg/ha of K2O/ha as potassium chloride and 50 kg/ha of P2O5 as rock phosphate. A completely randomised block design, with six treatments and five replications, was used.
Results
Rice yields, after incorporation of stylo stubble, were higher (P < 0.05) than those obtained with the control treatment or equal to those obtained with the application of 50
12
Rice yield as an indicator of soil fertility after the incorporation of S. guianensis stubble
kg/ha of N, which indicates the potential of this legume to supply N to the soil. However,
the best results were obtained with the application of 100 kg/ha of N (Table 1).
Table 1. Effect of incorporating Stylosanthes guianensis (stylo) stubble and of applying different levels
of fertiliser on rice yield and number of tillers and spikelets on the Sara farm in Ucayali, Pucallpa, Peru.
Treatments
Rice yield
(kg/ha)
Tillers
(no./m2)
Spikelets
(no./m2)
T1 = control (no treatment)
380c*
102a
136b
T2 = after incorporation of stylo
520b
111a
154ab
T3 = incorporation of stylo + 50 kg/ha of N
600b
119a
154ab
830a
125a
174ab
870a
130a
187a
890a
136a
185a
* Values in the same column followed by the same letters do not differ significantly (P < 0.05),
Study results indicate that, in the Pucallpa region, S. guianensis has a high initial
capacity for establishment and competition with weeds torourco. However, once
animals are introduced, its persistence tends to decrease, possibly because of its
superficial growth points or because of reduced soil fertility.
Conclusions
In Pucallpa, Peru, S. guianensis is a well-adapted legume with the capacity to fix up to 50
kg/ha of N when established as a pioneer crop for rice and subsequent establishment of
improved pastures. Its persistence is, however, low once the pasture is left for grazing.
13
Research results
Component 1
Activity 1.2 Use of forage resources and supplementation

of cows during the dry season
Highlights
The response of cows in terms of milk yield to forage legume supplementation, either
fresh or as silage, and sugarcane is similar to that obtained with protein sources from
commercial feed concentrates.
Direct grazing of shrub legumes increases milk yields as compared with the cut-andcarry system.
Legumes supply protein to forage basal diets that are usually nitrogen deficient.
Highly fermentable carbohydrates can stimulate the degradation of less fermentable
fibre.
The use of additives in the form of carbohydrates improves the quality of silage from
shrub legumes.
Evaluation of milk production systems using

Cratylia argentea
C. Lascano, P. Toro and P. Avila
CIAT, Colombia
Dual-purpose livestock production accounts for 78% of the livestock and 41% of the
milk produced in tropical Latin America. This type of exploitation is mainly found on
smallholdings (Rivas 1992) that depend, to a large extent, on forage resources that
present nutritional limitations and therefore restrict animal productivity.
Forage legumes are an alternative to improve animal nutrition in dual-purpose systems. This is true in the case of the shrub legume Cratylia argentea, a species that adapts
well to acid soils and preserves most of its leaves during the dry season. However, the use
of cut-and-carry Cratylia in some areas of the tropics is not viable because of the high
labour costs involved. This study aimed to evaluate the effect of Cratylia supplementation to cows in both cut-and-carry system and under direct grazing on Brachiaria
decumbens pastures.

The study was conducted during two contrasting seasons, at the CIAT experiment
station in Santander de Quilichao (Cauca, Colombia), located at an altititude of 990
metres above sea level (masl), with annual rainfall averaging 1700 mm. Trial 1 was
carried out from August to October 1999, a period corresponding to minimum
precipitation (water balance <50 mm). Trial 2 was carried out from May to June 2000
during maximum precipitation (water balance >50 mm).
Trials were arranged in a Latin square design (3 3) with the following treatments:
T1, B. decumbens associated with C. argentea under direct grazing
T2, B. decumbens with access to C. argentea at 1.5% DM liveweight (cut-and-carry) and
T3, B. decumbens alone, as control plot.
Each treatment was assigned to a 1-ha grazing area, divided into two equal sections to
allow 7 days for adjustment and 7 days for measurement. The animals spent 14 days in
each treatment for a total of 42 days. The stocking rate for each trial was two cows/
hectare; six crossbred Holstein Zebu cows with similar lactation status (secondthird)
and milk production levels was used.
17
Results
Availability of grass forage at the beginning of grazing was greater in Trial 2 as compared
with Trial 1. This difference in forage on offer was associated with changes in precipitation.
The quality of the grass varied considerably among treatments, the IVDMD and CP
levels being lower in the control treatment (B. decumbens alone) in both experiments. The
treatment and precipitation significantly affected both amount and quality of the grass.
Milk production of cows on the associated Cratylia + Brachiaria pasture under direct
grazing was consistently higher than that of the check treatment in both minimum (23%
more milk) and maximum precipitation (16% more milk), although this difference was
not significant (Table 1). The significant increase in milk production in Trial 1 can be
partially explained by the low amount and quality of the basal pasture, which was affected by the lower precipitation. These results confirm once more the importance of providing supplementation not only when the basal pasture is deficient but also during the
dry season, but not so in the rainy season (Avila 1999).
Table 1. Milk production of crossbred Holstein Zebu cows, supplemented with Cratylia argentea
in a cut-and-carry system as compared with direct grazing (Quilichao, Cauca, Colombia).
Treatment
Trial 1
Trial 2
Milk yield
(kg/cow per day)
Milk yield
(kg/cow per day)
Brachiaria decumbens
6.1b*
6.3a
B. decumbens + C. argentea (cut-and-carry)
6.7b
6.6a
B. decumbens + C. argentea (direct grazing)
7.5a
7.3a
* Values in the same column followed by the same letter do not differ significantly (P < 0.05), according
to Duncans test.
However, the levels of fat and non-fat solids (NFS) in milk were similar for both trials.
The milk urea nitrogen (MUN) levels differed significantly (P < 0.05) among treatments
for both trials. When C. argentea was used as supplement, MUN levels were higher compared with the check treatment. The levels of these indicators without C. argentea supplementation were marginal (Table 2). MUN levels ranged from 9 to 10 mg/dL, suggesting a protein deficit (Hammond et al. 1994) and accounting for the lower milk
production of the group of cows that did not receive legume supplementation.
Conclusions
The partial information from these trials indicates that the use of C. argentea under direct
grazing increased milk production the most. Also, direct supplementation with the legume is more viable during the drier season than during the rainy season because the
economic response is greater when the basal pasture has limited nutrients during the dry
season.
18
Evaluation of milk production systems using Cratylia argentea
Table 2. Quality of milk produced by cows supplemented with Cratylia argentea in a cut-and-carry system and
under direct grazing (Quilichao, Cauca, Colombia).
Trial 1
a
Trial 2
b
Treatment
Fat
(%)
NFS
(%)
3.5a*
8.0a
9.8b
B. decumbens + C. argentea (cut-and-carry)
3.6a
8.0a
33.6a
3.7a
8.4a
11.6a
B. decumbens + C. argentea (direct grazing)
3.7a
8.2a
27.3a
3.7a
8.2a
12.5a
MUN
(mg/dL)
Fat
(%)
NFS
(%)
MUN
(mg/dL)
3.6a
8.2a
3.0b
a. NFS = non-fat solids.

b. MUN = Milk urea nitrogen.
* Values in the same column followed by the same letters do not differ significantly (P < 0.05), according to
Duncans test.
References
Avila P. 1999. Efecto de la suplementacin con especies forrajeras en la produccin de leche de
vacas en pastoreo. MSc Thesis. Universidad Nacional de Colombia, Palmira, Colombia.
Hammond A.C., Kunkle W.E. and Genho P.C. 1994. Use of blood urea nitrogen concentration
to determine time and level of protein supplementation in wintering cows. Professional Animal
Scientist 10:2431.
Rivas L. 1992. El sistema ganadero de doble propsito en Amrica tropical. Paper presented at
the international symposium on animal production alternatives and strategies. Universidad
Autnoma de Chapingo, Mexico.
19
Nutritional complementarity between

forage basal diets and legume supplements
W. Quionez, P. Avila and C. Lascano
CIAT, Colombia
Previous results from confined feeding trials suggested that when supplementing available forage sources on the farm to overcome nutrient deficiencies in ruminants, it is
important to synchronise feeding of the higher quality forage supplement with the basal
forage so that energy and protein are available concurrently. Also, synergism among
forages may vary not only with the types of forages fed but also with how they are fed (for
example, amount and frequency).

Eight African type wethers (24 kg LW, on average) were fed a low-quality basal grass diet,
and were randomly allocated to four supplementation treatments, arranged in a 4 4
Latin square design, consisting of sugarcane (60%) mixed with Cratylia argentea (40%).
Treatments were as follows:
T1, low level of supplementation (0.5% LW) fed once a day (am)
T2, high level of supplementation (1.0% LW) fed once a day (am)
T3, low level of supplementation (0.5% LW) fed twice a day (am + pm) and
T4, high level of supplementation (1% LW) fed twice a day (am + pm).
Measurements included quality of the basal diet, intake of supplements offered,
digestibility and N balance.
Results
The low-quality basal grass diet was low in CP (4.8%) and high in cell wall content (79%
NDF and 44% ADF). However, chopped sugarcane, fed as an energy supplement, was
low in CP (3.1%) but had low cell wall content (39% NDF and 24% ADF). The legume
(Cratylia leaves) fed had high CP (21%) and high cell wall contents (67% NDF and 37%
ADF). Thus the supplement fed was high in energy and medium in protein (10% dry
basis). Intake of the basal diet did not differ among treatments, but there were differences in supplement intake due to treatments. As expected, intake of sugarcane and
Cratylia tended to be higher when fed at the highest level. However, it is interesting to
20
Nutritional complementarity between forage basal diets and legume supplements
note that when supplements were offered at the high level, intake of sugarcane and
Cratylia increased with twice-a-day feeding compared with once-a-day feeding (Table 1).
Table 1. Effect of level and frequency of supplementation with sugar cane and Cratylia on intake and digestibility of
African wethers fed a low-quality basal grass diet (Quilichao, Cauca, Colombia).
Frequency and level of supplementation1
2
Parameter
am
0.5% LW
am
1% LW
am + pm3
0.5% LW
am + pm
1% LW
SE4
25.2
25.7
0.8
DM intake (g/kg LW
per day)
Basal diet (grass)
25.7
25.2
Sugarcane
2.5b*
3.3b
2.8b
4.0a
0.4
Cratylia
2.0c
3.1b
2.0c
3.6a
0.1
Digestibility (%)
IVDMD
53.0
53.0
55.4
56.6
1.7
NDF
54.1
54.3
57.7
57.1
1.8
ADF
51.6
50.7
54.3
53.1
1.9
1. 60% sugarcane + 40% Cratylia argentea (leaves).

2. Supplement fed once a day at 0.5% or 1% LW.
3. Supplement fed twice a day at 0.5% or 1% LW.
4. Standard error.
* Values in the same column followed by the same letters do not differ significantly (P < 0.05), according to Duncans
test.
These differences in intake of supplements were not reflected in significant changes

in DM or cell wall digestibility. Nevertheless, there was a tendency for higher digestibility
with twice a day feeding at both levels of supplementation.
As expected, N intake was highest with increased level of supplementation (Table 2);
however, intake was highest when wethers were fed twice a day. Because faecal and urinary N did not change with treatment, N retention was greater when sheep were given the
forage-based supplements at the high level and twice a day.
Conclusions
This study evidenced that the level and frequency of supplementary feeding had a significant effect on N utilisation by growing African-type wethers. Results indicate, however,
that feeding twice a day would only be justified when high levels of forage-based supplements are offered. When the high level (1% LW) of sugarcane and Cratylia was fed
twice a day, there was a 33% increase in N retention relative to feeding the same amount
of supplement once a day. This was not the case when the low level (0.5% LW) was offered.
21
Table 2. Effect of frequency and level of supplementation of sugarcane and Cratylia on nitrogen utilisation by
African wethers fed a low-quality basal diet (Quilichao, Cauca, Colombia).
Frequency and level of supplementation1
am2
0.5% LW
am
1% LW
am + pm3
0.5% LW
N intake (g/day)
5.6a
6.2b
Faecal N (g/day)
3.5
Parameter
Faecal N (% N intake)
Urine N (g/day)
Urine N (% N intake)
Retained N (g/day)
am + pm
1% LW
SE4
5.6a
6.7c
0.01
3.5
3.3
3.5
0.1
62.5a
57.1b
59.9a
52.0c
2.0
1.4
1.5
1.5
1.6
0.2
25.4
25.8
28.6
24.7
3.7
0.8e
1.2d
0.8e
1.6a
0.2
1. Sixty percent sugarcane + 40% Cratylia argentea (leaves).

2. Supplement fed once a day at 0.5 or 1% LW.
3. Supplement fed twice a day at 0.5 or 1% LW.
4. Standard error.
test.
22
Effects of dry season feeding of fresh

and ensiled Cratylia argentea on milk
production and composition
ECAG, Costa Rica
One of the strategies of the Tropileche Consortium in Costa Rica is the search for feeding alternatives for cows during the dry season. The results of previous work carried out
by the Escuela Centroamericana de Ganadera (ECAG) indicate that sugarcane is an
alternative energy source for cows under grazing. Its protein content, however, is low.
The study aimed to evaluate the use of fresh and ensiled Cratylia argentea as a protein
source. The study was based on the hypothesis that the protein quality of fresh and
ensiled C. argentea was similar to that of yellow maize and soybean flour, which are
sources of protein for commercial concentrates.

The experiment was carried out between 25 February and 20 April 1999 at ECAG,
located in Balsa de Atenas, Costa Rica, at an altitude of 460 metre above sea level (masl),
with a mean annual temperature of 28C. Average precipitation was 1500 mm/year,
with two well-defined climatic seasons: a dry season from December to May and a rainy
season from June to November.
Forage management
Silage
The heap silage method was used. Silage was made during November 1998, using a
6-month-old Cratylia regrowth, which was cut to a height of 30 cm above the ground.
Usable parts (leaves and tender stems) were separated and chopped to a size of 2.5 cm.
Chopped material was then spread out in 20-cm layers and compacted. Molasses was
then added at a proportion of 10:1 (10 kg C. argentea:1 kg molasses).
23
Fresh Cratylia argentea

A 3-month-old regrowth was used. Plants were cut every day to a height of 30 cm above
the ground, then chopped and offered to cows mixed with sugarcane.
Sugarcane
Whole sugarcane was offered chopped (2.5-cm size) on a daily basis. Its nutritional content was 2% CP and 3.0 Mcal ME.
Concentrate
This concentrate was based on maize and soybean and prepared at the ECAG feed plant.
Its nutritional content was 14% CP and 2.3 Mcal ME.
Semolina
Commercial rice semolina, with 12% CP and 2.8 Mcal ME, was used.
Animal management
Six multifarious Jersey cows, at 50 days postpartum, were assigned at random to three
treatments in a Latin square design repeated simultaneously. Treatments and amounts of
products (in parenthesis as % LW) were as follows:
T1, sugarcane (1%) + semolina (0.5%) + concentrate (1.48%) + urea (0.02%)
T2, sugarcane (1.3%) + concentrate (0.5%) + fresh C. argentea (1.2%) and
T3, sugarcane (0.1%) + concentrate (0.5%) + silage of C. argentea (2.4%).
Measurements
Because of the limited amount of silage, each period covered 12 days, 7 of which corresponded to animal adaptation and to sampling. Samplings were made on days 1, 3 and
5.
Intake and quality of DM on offer and rejected

DM intake was calculated as the difference between the amount of material on offer and
that rejected on days 1, 3 and 5 during each five-day evaluation period. Composite
samples for that week were sent to the ECAG laboratory to determine crude protein
(CP) and in vitro dry matter digestibility (IVDMD).
24
Effects of dry season feeding of fresh and ensiled C. argentea on milk production
Milk production and quality

During the five days of each evaluation period, individual milk production (kg/cow) at
both milking was weighed.
To measure milk quality, two milkings proportional to production were sampled.
Each sample was cooled for conservation and then analysed for fat, protein, lactose, total
solids (TS) and non-fat solids (NFS).
Results
No major difference was observed among the three rations offered regarding total DM
intake. In the case of fresh C. argental, intake was 10.7 kg/cow per day (3.0% animal
LW) and that of silage was 10.4 kg/cow per day (2.9% animal LW), indicating that
Cratylia does not present problems of consumption when offered fresh or as silage
together with sugarcane (Table 1).
Table 1. Daily intake of dry matter (DM), crude protein (CP) and metabolisable energy (ME)
by cows offered different feeding treatments (ECAG, Costa Rica).
Daily intake
Treatment
DM
(kg)
DM
(% LW)
Concentrate
10.8

Ensiled C. argentea
Significance
CP
(kg)
ME
(Mcal)
3.1
1.3
24.2
10.7
3.0
1.3
23.6
10.4
2.9
1.4
21.9
0.5
0.1470
0.5924
0.5304
Table 2 indicates that the different treatments did not affect milk production of cows
significantly. These results do not coincide with those reported by Argel et al. (1999),
where milk production differed depending on whether fresh or ensiled C. argentea was
used (5.5 vs. 5.1 kg/cow per day).
Table 2. Production and composition of milk from cows offered different feeding treatments (ECAG, Costa Rica).
Treatment
Milk
(kg/cow per day)
Fat
(%)
Protein
(%)
Lactose
(%)
TS
(%)
NFS
(%)
Concentrate
11.1
3.53
3.36
4.80
12.39
8.86
10.9
3.69
3.24
4.84
12.47
8.78
Ensiled C. argentea
10.7
3.81
3.22
4.76
12.49
8.68
0.063
0.014
0.353
Significance
0.268
0.736
0.095
Except for protein, the values of all other milk components did not vary among treatments (Table 2). However, fat content tended to be slightly higher (P < 0.0627) in treatments that included C. argentea silage, which could benefit producers selling cream or
milk, when the latter is paid based on the percentage of fat.
25
But, most importantly, the milk production levels achieved with rations of C. argentea,
either fresh or ensiled (10.9 and 10.7 kg/cow per day, respectively), were comparable to
those obtained with a typical concentrate for dairy cows (11.1 kg/cow per day), in which
the protein comes from maize and soybean. The production levels reached with C. argentea
during the dry season would be fairly acceptable for medium and small producers. Furthermore, this legume could replace sources of energy and protein (maize and soybeans) used
to manufacture concentrates that are usually beyond their reach.
Fresh C. argentea with 3 months regrowth presents, on average, 19.9% CP and 53.4%
IVDMD (Table 3). This last value agrees with that reported by Lascano (1995), but differs from those found by Perdomo (1991) and Xavier and Carvalho (1995). In ensiled C.
argentea, the average values of CP (14.7%) and IVDMD (40.6%) were slightly lower than
those of fresh Cratylia; this level of PC, however, fulfils the requirements of the animals
used in the trial.
Table 3. Nutritional composition of fresh and ensiled Cratylia argentea (ECAG, Costa Rica).
Ensiled Cratylia argentea
Evaluation
periods
CP
(%)
IVDMD
(%)
CP
(%)
IVDMD
(%)
19.2
56.0
15.3
40.4
II
19.1
50.2
15.2
40.3
III
21.4
54.1
13.6
41.1
Average
19.9
53.4
14.7
40.6
Production costs
Costs were calculated based on the costs appearing in Table 4. The biomass production
of Cratylia at three months regrowth when plants reach their best leaf:stem ratio and
produce larger amounts of edible material, thus reducing labour costs.
Table 4. Feeding costs and milk components in cows fed concentrate and two forms of Cratylia argentea at three
months regrowth (ECAG, Costa Rica).
Milk components
(%)
Cost
(US$/kg)
Fat
Protein
Lactose
TS
NFS
Concentrate
0.21
5.85
6.15
4.30
1.67
2.33
Fresh Cratylia
0.17
4.56
5.19
3.48
1.35
1.92
Ensiled Cratylia
0.18
4.63
5.48
3.71
1.41
2.03
Treatment
TS = Total solids.
NSF = Non-fat solids.
Calculations considered the following costs per kg product on offer: concentrate

(processed at the ECAG plant), semolina, sugarcane, fresh C. argentea (taking into account cutting, carrying, and chopping) and ensiled C. argentea (preparation).
26
Effects of dry season feeding of fresh and ensiled C. argentea on milk production
Table 5 shows that, although gross income was higher when the concentrate was used
(US$ 3.13) compared with fresh or ensiled Cratylia (US$ 3.10 and 3.05, respectively),
costs were also higher and the net income was US$ 0.84. When fresh Cratylia was used
in the diet, net income was US$ 1.26 and with ensiled Cratylia it was US$ 1.16.
Table 5. Daily income and expenditures (US$/cow per day)
obtained with test diets (ECAG, Costa Rica).
Treatment
Income
Expenditures
Difference
Concentrate
3.13
2.30
0.84
Fresh Cratylia
3.10
1.83
1.26
Ensiled Cratylia
3.05
1.89
1.16
In addition to the economic benefit, the use of the legume is environmentally beneficial in terms of nutrient recycling and sustainability of the system.
Conclusions
The results of this study indicated that:
1. Fresh C. argentea, at three months regrowth, is a good alternative to feed milking cows
2. Under ECAG conditions, Cratylia should be harvested at three months regrowth to
prepare silage because the leaf:stem ratio is best then and
3. Fresh or ensiled C. argentea can replace the use of a commercial concentrate in dairy
cows of intermediate milk production during the dry season.
References
Argel P.J., Lobo M.B., Romero F., Gonzlez J., Lascano C.L., Kerridge P.C. and Holmann F.
1999. The shrub Cratylia argentea as a dry season feeding alternative in Costa Rica. Paper
presented at the workshop working with farmers: The key to adoption of forage technologies,
held in Cagayan de Oro, Northern Mindanao, Philippines, 1215 October 1999. 6 pp.
Lascano C.E. 1995. Calidad nutritiva y utilizacin de Cratylia argentea. In: Pizarro E.A. and
Coradin L. (eds), Potencial del gnero Cratylia como leguminosa forrajera. Proceedings of the
Cratylia workshop held 1920 July 1995. Brasilia, DF, Brazil. EMBRAPA, CENARGEN, CPAC,
CIAT (Centro Internacional de Agricultura Tropical), Cali, Colombia. pp. 8397.
Perdomo P. 1991. Adaptacin edfica y valor nutritivo de 25 especies y accesiones de leguminosas arbreas
y arbustivas en dos suelos contrastantes. Zootechny Thesis, Faculty of Agricultural Sciences,
Universidad Nacional de Colombia, Palmira, Colombia. 128 pp.
Xavier D.F. and Carvalho M.M. 1995. Avaliao agronmica da Cratylia argentea na zona da Mata
de Minas Gerais. Proceedings of the Cratylia workshop, held 1920 July 1995. Brasilia, DF, Brazil.
EMBRAPA, CENARGEN, CPAC, CIAT (Centro Internacional de Agricultura Tropical),
Cali, Colombia. pp. 2939.
27
Substitution of chicken manure

with Cratylia argentea in diets of cows
grazing Hyparrhenia rufa pastures
M. Ibrahim, M. Franco, D. Pezo, A. Camero and J. Araya
CATIE and MAG, Costa Rica
The production and quality of forage biomass from grasses is drastically reduced in ecosystems with prolonged droughts, such as the Central Pacific region of Costa Rica. Producers solve this limitation by supplementing with agricultural by-products, such as
chicken manure. Several years ago this by-product was inexpensive but its price in real
terms has increased in view of the growing demand. Producers are therefore interested in
evaluating alternatives to replace chicken manure. The results of three on-farm trials are
reported: two trials evaluated the substitution of chicken manure with Cratylia argentea
and the other trial evaluated cratylia fed fresh or as silage.

Experiment 1
This trial was conducted during the dry season (April 1999) on a small farm in Miramar,
located at an altitude of 250 metres above sea level (masl) in the Central Pacific region of
Costa Rica. The mean annual temperature is 28C with a mean precipitation of 2400
mm. The following treatments were evaluated:
T1, control, only grazing (naturalised grass and jaragua grass)
T2, 12 kg sugarcane + 8 kg C. argentea + 0.6 kg rice polishing + grazing and
T3, 12 kg sugarcane + 3 kg chicken manure + 0.6 rice polishing + grazing.
Cows in T2 and T3 received molasses.
The nutritional characteristics of the feed were as follows: sugarcane, 2% CP and 3.0
Mcal ME; chicken manure, 19.5% CP and 1.8 Mcal ME; rice polishing, 12% CP and 3.0
Mcal ME; fresh cratylia, 20% CP and 1.8 Mcal ME (from 90-day regrowth cut at 30 cm);
and cratylia silage, 16.4% CP and 1.9 Mcal ME (from 180-day regrowth cut at 30 cm).
A Latin square design with three treatments was used, each treatment with two selected cows in the second month of lactation. The experiment lasted for 30 days, and
animals were rotated through each treatment at 10 day intervals (7 days for adjustment
and 3 for data collection).
28
Experiment 2
This trial was conducted in a small farm in Barranca, located at an altitude of 280 masl
in the Central Pacific region of Costa Rica. The mean annual temperature is 28C and
mean precipitation of 2500 mm. Treatments evaluated were:
T1, 12 kg sugarcane + 6 kg silage of C. argentea + 0.6 kg rice polishing
T2, 12 kg sugarcane + 6 kg fresh C. argentea + 0.6 kg rice polishing and
T3, 12 kg sugarcane + 3 kg chicken manure + 0.6 kg rice polishing.
A Latin square design with three treatments was used, each treatment with two cows
as in the above experiment. Animals were managed in individual groups. The experiment lasted for 30 days and animals were rotated on each treatment at 10-day-intervals
(7 days for adjustment and 3 for data collection).
Experiment 3
This on-farm trial was conducted as part of a student thesis programme on a small farm
in Barranca, located at an altitude of 280 masl in the Central Pacific region of Costa
Rica. The trial was conducted in the middle of the dry season (FebruaryMay 1998),
with a mean annual temperature of 28C and a mean precipitation of 2500 mm.
Treatments evaluated were:
T1, chicken manure and molasses
T2, chicken manure, sugarcane, molasses and wheat bran
T3, chicken manure, molasses, C. argentea and wheat bran.
Details of the different diets are presented in Table 1. Different amounts of supplements were used to balance the diets isonitrogenously and isocalorificly. C. argentea
satisfied more than 75% of the CP requirements of animals, but a small amount of
chicken manure was included in the diet because observed C. argentea intake was not
sufficient to balance N requirements. Animals had access to H. rufa pastures. Cratylia CP
content was 19.2% with 58.4% IVDMD; the respective values for H. rufa were 3.9% and
33.9%.
A 3 3 Latin square changeover design with three replications was used with a total
of 9 crossbred cows between 6080 days into lactation. Each of the three experimental
periods consisted of 10 days for adaptation to treatment and 5 days for collection of experimental data.
Results
Experiment 1
Table 2 shows milk yield and composition, feeding costs, income from milk, and the
cost:benefit ratio for the different treatments evaluated. Milk yields of T2 and T3 did
not differ significantly (P = 0.076), indicating that C. argentea can be used to substitute
29
Substitution of chicken manure with C. argentea in diets of cows grazing H. rufa pastures
chicken manure without reducing milk production. Furthermore, feeding costs are lower
and the cost:benefit ratio is higher when C. argentea is used, which makes this alternative
more economically attractive to farmers.
Table 1. Diets offered to animals and milk yields for all treatments.
Diets
Consumption
(kg/animal per day)
Milk yield
(kg/cow per day)
Diet 1
5.9
Chicken manure
6.0
Molasses
2.5
Diet 2
6.0
Chicken manure
5.0
Sugarcane
5.0
Wheat bran
0.7
Molasses
0.12
Diet 3
6.1
Chicken manure
1.0
Molasses
4.12
Wheat bran
0.7
Cratylia argentea
6.0
Table 2. Average milk production and composition, cost of supplementary feeding (Cratylia argentea and chicken
manure), income from milk, and cost:benefit ratio of diets offered to dual-purpose cows.
Treatments
Milk yield
(kg/cow per
day)
Total
solids
(%)
Fat
(%)
Cost of
supplementary
feeding
(US$/kg milk)
Income
(US$/kg
milk)
Cost:
benefit
ratio
Only grazing
5.45b*
11.2
3.1
0.24
Grazing + Cratylia
5.85ab
11.5
3.2
0.11
0.25
2.22
Grazing + chicken manure
6.29a
11.2
2.9
0.22
0.24
1.08
test.
Experiment 2
Table 3 shows that the feeding alternatives based on C. argentea, both ensiled and fresh,
were more economical than the one that used chicken manure as supplement. Not only
are Cratylia rations more economical, the milk yields of cows consuming Cratylia, either
fresh or ensiled, were similar to those of cows supplemented with chicken manure. However, milk yields of cows consuming Cratylia silage were lower than those of cows consuming fresh Cratylia. Nevertheless, Cratylia silage can be used as a substitute for chicken
manure because of the lower feed costs.
30
Table 3. Average milk production and composition, cost of supplementary feeding (fresh or ensiled Cratylia argentea
and chicken manure), income from milk, and cost:benefit ratio of diets offered to dual-purpose cows.
Treatments
Milk yield
(kg/cow per
day)
Total
solids
(%)
Fat
(%)
Cost of supplementary
feeding
(US$/kg milk)
Income
(US$/kg
milk)
Cost:benefit
ratio
Ensiled Cratylia
5.09b*
12.3
3.7
0.17
0.27
1.58
Fresh Cratylia
5.47a
12.2
3.5
0.11
0.27
2.37
Chicken manure
5.26ab
11.7
3.0
0.22
0.26
1.14
test.
Experiment 3
Table 3 shows the effect on milk yield per treatment. Milk yield averaged 6.0 kg/cow per
day and there were no significant differences between treatments. Milk fat was lower
(2.7%) for the treatment that had a higher amount of sugarcane in the diet, but differences were not significant. The results clearly show that Cratylia can be used as a substitute for chicken manure, at least up to 82% of the diet.
Conclusions
In recent years, most farmers with dual-purpose cattle have produced milk during the dry
season using chicken manure. Results clearly show that Cratylia can largely replace chicken
manure as a protein supplement without significantly reducing milk yield. The growing
demand for chicken manure to supplement dairy cows has increased its price in many
tropical regions and farmers can no longer afford to buy it. By introducing C. argentea to
cattle production systems in the Pacific region, the use of chicken manure has decreased
in farms participating in the Tropileche Consortium to such degree that this year practically none was purchased. Therefore, the legume technology being promoted for dualpurpose cattle farms could allow small producers to have access to a farm-grown protein
supplement and increase their cash flow and profits. In addition, spontaneous adoption
of C. argentea is occurring in other locations in Costa Rica. During the last 12 months,
84 kg of experimental seed has been sold to 28 farmers located in three different
sitesGuanacaste, Nicoya and around Esparza.
The most economical option for a producer in the dry season is to supplement cows
with fresh Cratylia. The next best option is to supplement with ensiled Cratylia. The least
economical option is to supplement cows with chicken manure. As a result, the use of
legume silage such as C. argentea is recommended over the use of chicken manure. In
addition, the use of legume silage allows producers to use smaller areas because more
edible biomass is produced for dry season supplementation, especially in situations
where the opportunity cost for labour during the rainy season is low.
31
Use of additives in Cratylia argentea silage

C. Jimnez, L. Pineda and A. Medina
UCR, Costa Rica
In the dry tropics of Costa Rica (0800 metres above sea level (masl), 12001800 mm of
precipitation; 6 months of dry season), Cratylia argentea has been successfully introduced
into dual-purpose farms to solve the problems of low forage availability during summer
months. Mixing cratylia forage and chopped sugarcane has proved useful to maintain
milk yields of 66.5 kg/cow per day in crossbred Zebu European dairy cows (Argel and
Lascano 1998, Lobo and Acua 1998).
Producers value cratylias high CP content (more than 15%) and its adaptability and
persistence. Furthermore, one of its advantages is that surplus forage can be ensiled for
use during times of feed shortage. This is especially important because cows do not
readily consume cratylia forage when the offer of other forage grasses in the pastures is
high.
This study aims to evaluate the process of making silage from cratylia, using several
fermentative and nutritive additives. Overall, legumes are more difficult to ensile than
grasses because they have a high buffer capacity that hinders adequate acidification of
anaerobic media, in addition to its low soluble carbohydrate content.

The foliage of 90-day legume regrowth was manually harvested and chopped to a 16 mm
bite size. Microsilos made of double polyethylene plastic bags, with approximately 1.5 kg
capacity, were used. Cratylia was mixed with three additives: cane molasses (M), pineapple pulp (PP) and chopped sugarcane (CSC). Each additive was incorporated at three
levels: M at 10, 20 and 30%; PP and CSC at 25, 50 and 75%, all on fresh basis (w/w).
Treatments were arranged in a completely randomised design with three replicates.
Cratylia foliage was obtained from an experimental plot at the Escuela Centro-americana de Ganadera (ECAG), located in Atenas, Costa Rica, at 460 masl. Pineapple pulp
was collected at the Del Oro plant in Santa Cecilia de la Cruz, Guanacaste and the sugarcane was gathered at a private farm located in the Hojancha canton, also in Guanacaste.
The silos were left to ferment for 60 days and, upon opening, pH and organoleptic
characteristics (odour and colour) were assessed. Part of each sample was oven-dried at
60C and a sub-sample was frozen to later analyse ammonium nitrogen. Laboratory tests
were conducted to determine dry matter (DM), organic matter (OM), crude protein
(CP), ammonium nitrogen (Nam) and rumen degradability of DM (D).
32
Use of additives in Cratylia argentea silage
A scale (from 1 to 3) was used for the organoleptic evaluation.

Scores and properties of silage
Organoleptic
characteristics
1 = Poor
2 = Intermediate
3 = Good
Odour
Butyric (rancid)
Acetic
Lactic
Degraded amino acids or N

sub-products (spoiled)
Alcoholic
Slightly acetic or alcoholic

with lactic essence
Dark brown
Black
Mouldy
Light brown
Yellowish
Original green
Light green
Colour
Results and discussion

All silages presented a strong lactic odour, except for those with high CSC content
(75%), which presented an alcoholic odour with a lactic essence (Table 1). The silage
with 25% PP and high levels of CSC presented a light brown colour.
Table 1. Characteristics* of Cratylia argentea silage using three types of additives.
Additive
Ratio
(w/w, %)
Odour
Molasses
10
4.12
20
4.00
30
4.05
25
2.33
4.45
50
3.90
Pineapple pulp
Chopped sugarcane
Colour
pH
75
2.83
3.45
25
2.5
3.37
50
2.5
4.00
75
2.83
2.33
4.73
* Silage quality: 1 = poor, 2 = intermediate, 3 = good.
The pH correlated highly with lactic and acetic acid contents. A pH lower than four
is considered a good indicator of superior fermentation in moist substrata such as the
ones used in this project and, accordingly, of a good-quality end product. When carbohydrate levels are very high or very low, fermentation is altered, yielding final products
and pH values that fall outside the optimal range. In the case of 75% CSC, the high
levels of soluble carbohydrates (SC) induced fermentation tending more toward an
alcoholic base, with a pH of almost 5. In the case of 25% PP, because of the high moisture content of this substratum, its SC contribution (dry basis) was small, causing lactic
acid production to be poor. Therefore pH was almost 4.5.
The buffer effect was evident in Cratylia, especially in the treatment with 10%
molasses. In elephant grass (Pennisetum purpureum), the addition of 10% cane molasses
33
increases the SC content to more than 11%. Fermentation is also optimal with pH
values close to 3.5 (Vargas et al. 1981; Chacn 1987). Furthermore, buffer capacity could
limit carbohydrate use and lactic fermentation and, accordingly, high pH values in
several other treatments, especially those with higher/lower SC contents.
Conclusions
Based on available data, the following conclusions and recommendations were made:
The addition of 10% molasses should be sufficient to obtain good cratylia silage.
Pineapple pulp added at levels higher than 25% improves fermentation of cratylia
silage, although this material has a high water content that increases transportation
costs.
Chopped sugarcane, added at 25%, contributes SC for good lactic fermentation; at
higher levels, there is a risk of promoting an alcoholic process.
References
Argel P.J. and Lascano C.E. 1998. Cratylia argentea (Desvaux) O. Kuntze: Una nueva leguminosa
arbustiva para sulos cidos en zonas subhmedas tropicales. Pasturas Tropicales 20(1):3743.
Chacn U.H. 1987. Determinacin de los cambios fsicoqumicos durante la fermentacin del pasto
elefante (Pennisetum purpureum) en microsilos. Thesis Agronomy, School of Zootechny, Faculty
of Agronomy, University of Costa Rica, Costa Rica. 72 pp.
Lobo M. and Acua V. 1998. Uso de Cratylia argentea y caa de azcar como un sustituto de la
gallinaza en fincas de doble propsito en las laderas secas de Costa Rica. Tropileche Hoja
Informativa 5, October 1998.
Vargas C.R., Jimenez C., Boschini C. and Constela M. 1981. Estudio sobre cambios fsicoqumicos durante la fermentacin del pasto elefante en microsilos de laboratorio con tres
niveles de melaza. Agronoma Costarricense 5(1/2):121125.
34
Planting maize in association with forage

soybean for silage production
I. Economic evaluation of partial sale of tender
corn harvest
UCR, Costa Rica
In the tropics no other forage is superior to maize as silage. However, high production costs
are the main constraint faced by both small- and medium-scale producers in Latin America.
In recent years, beef cattle herds are quickly becoming dual-purpose herds along
Costa Ricas North Pacific Coast. The milk produced by these farms is marketed as
cheese, sold fresh to consumers or delivered to industrial co-operatives. Except for the
milk destined to the co-operatives, summer prices (December to May) are higher because
of the low supply and steady demand.
To increase the availability of milk, the shortage and poor quality of forage on offer
during the dry season must be corrected. Some dairy cattle farms in the region are using
sugarcane and feed supplements, for example chicken manure, urea and molasses. These
products, however, do not meet the requirements of genetically improved cows, which
need forages of higher nutritive value to allow animals to best express this potential
during the dry season.
The Alfredo Volio Mata dairy cattle experiment station and the School of Zootechny
of the University of Costa Rica, with the collaboration of the Ministry of Agriculture
and Animal Husbandry of Costa Rica (MAG), are carrying out a project to evaluate new
feeding technologies on 16 farms with genetically improved cows. During project execution new challenges have arisen that need to be resolved, for example, the lack of
better-quality forages to meet animal requirements during the dry season.
This study aimed to evaluate the factors that intervene in the production of maize
and soybean silage, with special emphasis on the costs involved in cultivation and silage
making.

In August 1999 a locally adapted white maize hybrid (Cristiani Burkard HS-5G) and a
soybean variety for grain and forage (CIGRAS 10), developed by the Grain and Seed
Research Centre of the University of Costa Rica and successfully tested in the region,
were planted. Maize was planted at a rate of 6 seeds/m of furrow and soybean at 15
35
seeds/m of furrow. In both cases, distance between furrows was 0.80 m. Main plots consisted of furrows of each crop as follows:
1. associated (1:1)
2. intercropped (4:4) and
3. monoculture.
A split-split plot design was used in which subplots consisted of three forms of crop
harvesting before silaging:
1. removal of 50% of the baby corn (at mid-flowering)
2. removal of 50% of mature maize (tender grain) and
3. removal of nothing.
Experiment variables consisted of planting method and percentage of removal at harvest of mature maize or baby corn. Each treatment was replicated three times.
Maize was planted in association to improve CP content of the final product. Modifications at harvest were included to estimate the economic impact of selling part of the
agricultural harvest and the capacity to recover either partially or totally the high investment made in the crop and silaging process.
The yields of baby corn, mature maize or tender grain, and forage of each crop were
determined and the detailed costs of maize cultivation and silage making.

Table 1 details the costs per hectare for planting, maintenance, harvest and silage making,
assuming equal areas of each associated crop. Table 2 summarises crop yields for all treatments, forage being expressed in both fresh and dry basis.
The average sale of baby corn was 7533 units/ha, at US$ 0.02 each, totalling US$
150.66. If costs of growing maize and making silage are deducted from the harvest of
baby corn, their sale only recovers 37.5% of total costs/ha. In the case of mature maize,
the average harvest was 7312 units at US$ 0.0425/unit, representing a total income of
US$ 310.76, which covered 77.4% of the costs.
The average yield in terms of maize ears per plant was less than 1, which is much
lower than that expected, significantly affecting results. With acceptable plant density
(from 4 to 5 plants/m of furrow) and with 1 ear or more per plant, 50% of baby corn or
mature maize can be harvested, which would represent from 25 to 30 thousand units/
ha.
The maize produced approximately 30 t fresh forage/ha and soybean, 11 t/ha, which
are equivalent to 10.5 and 3.5 t DM/ha. In the case of maize, optimal yields are between
40 and 45 t/ha. These yields can be improved by applying a more balanced fertilisation
and by correcting the poor drainage in some plots. Soybean yields were 50% of those
obtained in monoculture in the Carrillo canton (Costa Rica), in very similar soil and
climatic conditions.
36
Planting maize in association with forage soybean for silage production
Table 1. Production costs of maize for silage making (Pacific Coast, Costa Rica, 1999).
Cost component
Units/ha
Unitary cost
(US$)
Total cost
(US$)
11.7
46.8
Planting
Maize seed (kg)
Soybean seed (kg)
Inoculant (kg)
Fertiliser (10:30:10) (kg)
Labour (h)
18
16
0.32
100
36
0.85
0.4
13.35
0.3
1.09
15.3
6.4
4.27
30
39.24
Crop maintenance
Fertiliser (urea, kg)
Herbicide (atrazine, L)
Labour (h)
300
2
12
0.2
5
1.09
60
10
13.08
Harvest
Baby corn (h)
Mature maize (h)
Forage (cut-and-carry) (h)
Stationary chopper (t)
14
14
40
25
1.09
1.09
1.09
0.17
15.26
15.26
76
4.25
Silage (heap silo)

Plastic (t)
Compaction (t)
Labour (h)
25
25
25
0.42
1.12
1.09
10.5
28
27.25
401.61
Land preparation (h)
Total (US$)
Conclusions
Mixed cropping of maize and soybean for silage making has great potential in the Pacific
region of Costa Rica, not only for increasing the amount of protein in the final product,
but also because of the economic effect on the final value of the product and partial use
of the maize harvest. Results show that it was possible to pay 77.5% of the cost of making
silage, including planting, harvest and maintenance.
In this case, if cropping conditions are improved, especially in terms of appropriate
fertilisation and improved soil drainage in some plots, then yields will improve and more
mature maize can be sold, allowing 100% or more of the initial investment to be recovered.
37
Table 2. Yields per hectare in different methods of cultivating maize and soybean for silage making (Pacific Coast,
Costa Rica, 1999).
Forage
Production
(No./hectare)
Treatment
BC
MM
Fresh (kg/ha)
2
Maize
Soybean
Dry (kg/ha)
Maize
Soybean
In association
Harvest of 50% BC
Harvest of 50% MM
No harvest
14,780
0
0
0
15,870
0
28,194
29,584
40,834
10,278
7084
11,458
9365.4 (74.9)
11,131.2 (82.9)
14,150.8 (79.0)
3014.6 (25.1)
2285.4 (17.1)
3679.0 (21.0)
Alternate
Harvest of 50% BC
Harvest of 50% MM
No harvest
14,890
0
0
0
13,358
0
26,528
24,930
33,958
13,820
12,362
14,270
9829.4 (70.4)
8999.8 (71.2)
12,267.4 (73.0)
4158.4 (29.6)
3597.0 (28.8)
4248.0 (23.0)
Monoculture
Harvest of 50% BC
Harvest of 50% MM
No harvest
15,533
0
0
0
14,645
0
31,944
24,306
30,972
10,556
10,278
11,216
9795.2 (76.8)
8448.4 (70.9)
11,386.8 (74.6)
3002.8 (23.2)
3239.4 (29.1)
3872.8 (25.4)
15,067
0
0
0
14,624
0
15,067
14,624
28,888.70
26,273.30
35,254.70
32,870.70
28,472.00
29,074.00
30,138.90
11,551.30
9908
12,314.70
9606.7
13,484.00
10,683.30
11,258.00
9663.3 (74.0)
9526.5 (75.0)
12,601.7 (75.5)
11,549.1 (78.6)
10,365.5 (71.5)
9876.8 (74.1)
10,597.2 (75.0)
3391.9 (26.0)
3040.6 (25.0)
3933.3 (24.5)
2993.0 (21.4)
4001.1 (28.5)
3371.7 (25.9)
3455.3 (25.0)
Averages
Harvest of 50% BC
Harvest of 50% MM
No harvest
In association
Alternate
Monoculture
General
1. BC = baby corn; 2. MM = mature maize.

Figures in parenthesis imply the percentage of each associated component in total dry matter.
38
Planting maize in association with forage

soybean for silage production
II. Indicators of silage quality
UCR, Costa Rica
The incorporation of a legume into the crop when making silage improves the quality of
the end product. Also, the sale of part of the baby corn in maize fields could reduce production costs.
On Costa Ricas North Pacific Coast, cattle owners have very sound dual-purpose
dairy farms but face problems to feed cows during the dry season. Under these conditions, the use of good-quality silagefor example, maize plus forage soybeanwould
allow these farmers to maintain milk production during this season.
The study was carried out at the Alfredo Volio Mata dairy cattle experiment station
and the School of Zootechny of the University of Costa Rica, and complements previous
studies. It aims to evaluate the quality of silage made from maize grown in association
with soybean.

Microsilos made of double polyethylene plastic bags with approximately 3 kg capacity
were prepared with recently harvested maize and forage soybean in different proportions
(Table 1). As in the previous study, the following maize forage conditions were considered:
removal of 50% of the baby corn (at mid-flowering)
removal of 50% of mature maize (tender grain) and
removal of nothing.
Maize was harvested one week after tender grain was harvested and soybean when it
reached phenological state R6 (seed formed and beginning to fill).
The microsilos were left open after 70 days of fermentation; and pH and organoleptic
characteristics (odour and colour) were assessed according to the scale in Table 2.
Part of each sample was oven-dried at 60C and a sub-sample was frozen to later
analyse ammonium nitrogen (Nam). Laboratory tests were conducted to determine:
1. dry matter (DM), organic matter (OM) and crude protein (CP) contents
2. Nam and
3. rate of rumen degradability of dry matter (D).
39
Table 1. Treatments evaluated for making maize + forage soybean silage

(Pacific Coast, Costa Rica).
Treatment
Maize:soybean ratio
Condition of forage
T1
1:1
No removal of ears
T2
1:1
Removal of 50% tender maize
T3
1:1
Removal of 50% baby corn
T4
2:1
No removal of ears
T5
2:1
T6
2:1
T7
3:1
No removal of ears
T8
3:1
T9
3:1
Table 2. Scales of organoleptic characteristics.

Organoleptic
characteristics
Scores and properties of silage

1 = Poor
2 = Intermediate
3 = Good
Odour
Butyric (rancid)
Degraded amino acids or N
sub-products (spoiled)
Acetic
Alcoholic
Lactic
Slightly acetic or alcoholic
with lactic essence
Colour
Dark brown
Black
Mouldy
Light brown
Yellowish
Original green
Light green
Treatments were organised in a completely randomised design with three replicates.

The organoleptic characteristics of the silage are indicated in Table 3. The odour ranged
from slightly acetic to lactic; as evidenced by the pH values obtained. These values correlated with lactic fermentation. As reference, three replicates of pure soybean were fermented and their average results were pH = 5, odour = 2, and colour = 3. The high pH
indicates the high buffer capacity of legumes that can affect final fermentation results,
even with as good a substratum as maize. The use of a lactobacillus as inoculums could
improve fermentation characteristics. All silos showed an excellent conservation of the
original colour.
Soybean undoubtedly improves CP content, reaching almost 12% in a mixture with
equal parts of maize:soybean (w/w) in fresh basis (Table 4). The CP content of similar
silage made only of soybean was 15.6%. According to data of the University of Wisconsin,
the average CP content of 1996 samples of maize silage, collected over 10 years in Iowa
and Wisconsin, was 7.6 0.81% (Lauer et al. 1999). Crude protein tends to decrease
with decreasing percentage of soybean in the mixture, reaching values close to 8%.
40
Planting maize in association with forage soybean for silage production
Therefore, the data obtained in this experiment agree with normal values for maize silage
(7.6 0.81%). The DM contents of the mixtures were very similar and the addition of
soybean had no effect, which is advantageous because it would not reduce animal DM
intake.
Table 3. Odour, colour and pH of maize:soybean silage in different proportions
(Pacific Region, Costa Rica).
Maize:soybean
ratio
Odour
Colour
pH
T1
1:1
2.50
3.00
4.25
T2
1:1
2.50
3.00
3.50
T3
1:1
3.00
3.00
3.95
T4
2:1
T5
2:1
2.50
3.00
4.00
T6
2:1
2.00
2.50
4.20
T7
3:1
2.50
3.00
3.80
T8
3:1
2.50
3.00
3.65
T9
3:1
2.75
3.00
3.68
T9
3:1
2.75
3.00
3.68
Treatment*
* Treatments are described in Table 1.
Table 4. Dry matter and crude protein contents of maize:soybean silages

in different proportions (Pacific Region, Costa Rica).
Maize:soybean ratio
DM
(%)
CP
(%)
T1
1:1
29.3
12.0
T2
1:1
26.5
11.6
T3
1:1
27.4
11.9
27.7
11.8a
Treatment*
Average**
T4
2:1
T5
2:1
26.9
9.8
T6
2:1
26.1
9.2
26.5
9.5b
Average
T7
3:1
26.9
8.1
T8
3:1
27.7
8.8
T9
3:1
30.2
8.7
28.3
8.5c
Average
* Treatments are described in Table 1.

** Averages followed by the same letter do not differ significantly
(P = 0.05%), according to Duncans test.
41
Titterton and Maasdorp (1997) worked with silage (1:1) of maize and 15 legumes in
monoculture. They found that all legumes presented acceptable levels of fermentation
(pH ranging between 3.7 and 4.5; NH3:N > 12%). They also found that protein content
(dry basis) increased from 77 g/kg in pure maize silage to 93 g/kg in yellow lupine and
153 g/kg in soybean forage.
Conclusions
Based on the organoleptic evaluation and pH values obtained, the silages obtained in
this study can be considered as of very good quality. DM contents, which ranged
between 26.5% and 28.3%, are acceptable and associated with a low loss of effluents.
Silage quality values are within the normal ranges cited in literature and clearly
evidence the advantage of using soybean forage in mixtures with maize for silaging.
References
Lauer J., Kohn K. and Flannery P. 1999. Wisconsin corn hybrid performance trials: Grain and silage.
University of Wisconsin, College of Agricultural and Life Sciences. Bulletin A3653.
December 1999.
Titterton M. and Maasdorp B. 1997. Nutritional improvement of maize silage for dairying: Mixed
crop silages from sole and intercropped legumes and a long season variety of maize. 2.
Ensilage. Journal of Feed Science Technology 69:263270.
42
Research results
Component 1
Activity 1.3 Functional relationships among forage

resources of contrasting quality, rumen fermentation
and blood parameters
Highlights
Milk urea nitrogen (MUN) concentrations can be a good indicator of the protein:
energy ratio in the diet.

MUN concentrations prove useful to increase the efficiency in use of on-farm feed
resources.
Milk urea nitrogen concentration

for recommendations on optimum proteinto-energy ratios in tropical forage diets
P. Avila and C. Lascano
CIAT, Colombia
When there is an excess of nitrogen relative to energy in the rumen, ruminal ammonia
concentration increases. Unused urinal ammonia enters the portal blood through the
rumen wall and is transferred to the liver where it is detoxified by conversion to urea.
The liver also produces urea from diminution of amino acids rising from postruminal
digestion and systemic protein turnover. Urea then circulates in the blood to the kidneys
and is excreted with the urine or it can be diffused from the blood into milk. When
there is a deficiency of dietary proteins, urinal ammonia concentrations are relatively low
and the proportion of nitrogen recycled back to the rumen as urea increases. As a result
of these metabolic transactions, blood urea nitrogen (BUN) is highly correlated with
urinal ammonia and milk urea nitrogen (MUN). Therefore, in healthy ruminants, MUN
concentrations could be a good indicator of the protein-to-energy ratio in the diet.
The purpose of the study was to determine the level of urea in the milk that could be
used as reference to increase the protein content to supplement diets, with a high probability that cows will respond by increasing milk production.

The study was carried out between 1992 and 1995 at the CIAT-Quilichao experiment
station on contrasting pastures of Andropogon gayanus and Brachiaria dictyoneura. Two
groups of four Brahman Holstein cows were submitted to grazing on each pasture, with a
7-day period to adjust cows to the diet and 7 days of measurements. Milk yields were
recorded during the 7-day measurement period, and milk samples for urea content were
taken on days 1, 4 and 7 of each period. A 4 4 Latin square experiment design was used.
Results
Figure 1 shows the relationship between milk yield increase and MUN of cows supplemented with Cratylia argentea and sugarcane. Four groups can be identified, one in
each quadrant. In the first group, cows showed an increase in milk yield when offered a
legume supplement when the level of urea in milk was <10 mg/dL. These observations
45
corresponded mainly to crossbred cows. In the second group, cows did not respond to
legume supplementation even when the urea level was <10 mg/dL. These observations
corresponded to crossbred and Brahman cows in similar proportions. The third group
included cows with urea level >10 mg/dL that showed a modest increase in milk production. This group was composed of both crossbred and Brahman cows in similar proportions. The fourth group included those cows who did not respond to legume supplementation with an increase in milk yield when urea level in milk >10 mg/dL. This
group was mostly composed of Brahman cows.
Milk production increase due to legume supplementation (%)

with respect to the control (sugarcane alone)
56
42
28
14
14
28
0
12
16
20
Urea in milk (mg/dL)
Figure 1. Relationship between milk yield increase and urea level in milk of grazing cows
supplemented with Cratylia argentea and sugarcane.
Conclusions
These results suggest a urea level in milk of about 10 mg/dL could be used as a benchmark figure to increase dietary protein content with a high probability that milking cows
will respond with increased milk yields provided demonstrated genetic potential.
Furthermore, these results suggest that crossbred cows respond to legume supplementation, even when MUN levels are above 10 mg/dL. This did not seem to be the case for
most Brahman- or Zebu-type cows.
46
Research results
Component 2 Developing new feeding systems
for dual-purpose cattle
Activity 2.1 On-farm evaluation of different forage
systems for beef cattle in dual-purpose systems
Highlights
Forage-based feeding systems for strategic supplementation during the dry season
reduce the need to purchase animal feed concentrates and increase the income in
dual-purpose farms.
The use of legumes with an energy source during the dry season maintains milk
production.
Milk production of dual-purpose cows

grazing Brachiaria brizantha cv. La Libertad
alone and associated with Arachis pintoi
cv. El Porvenir
M. Lobo and V. Acua
MAG, Costa Rica
Costa Rica has a land area of 51,023 km2, of which 25,100 km2 (49.2%) is under agricultural and livestock production according to the 1984 National Agricultural and Livestock Census. Of the area under cultivation, grasses occupy the largest proportion
16,500 km2 (65.7%). Beef cattle are raised on 51 thousand farms; their population is
estimated at 2.15 million head, of which 75% have been crossbred, at different levels,
with Zebu breeds.
Among the main problems affecting beef cattle development in Costa Rica are low
meat and milk productivity, low reproductive rates, nutritional deficiencies (especially
because of the low forage availability and quality during the dry season), and health
problems.
Forage is the most available and inexpensive source of feed for livestock in Costa
Rica; however, a high percentage of pastures are now degraded, reducing forage availability and quality.
The improvement of both the quality and quantity of forage resources through the
recovery of degraded pastures is important to increase overall livestock productivity. The
study aims to evaluate milk production of dual-purpose cows grazing a recovered pasture
of Brachiaria brizantha cv. La Libertad alone and associated with Arachis pintoi cv. El
Porvenir and Centrosema brasilianum.

The trial was carried out on a farm located in San Jernimo de Esparza, in the subhumid tropical central Pacific region of Costa Rica, at an average altitude of 250 metres
above sea level (masl). The region is characterised by having well-defined dry (December
May) and rainy (JuneNovember) seasons, with an average annual precipitation of 2500
mm. Four hectares of Brachiaria brizantha cv. La Libertad were planted in a paddock
where a degraded pasture of Brachiaria ruziziensis existed. The area was divided in half
and two hectares were planted in association with A. pintoi CIAT 18744 cv. Porvenir and
C. brasilianum CIAT 5234.
49
Forage availability (dry basis) and botanical composition of both pastures (alone and
associated) were determined using the Botanic method.
A group of 26 milking cows of different breedspredominantly Brown Swissand
weighing 450 kg on average were used; a Brahman bull was also included. This group
grazed the two hectares of sole cropped pasture for four consecutive days and the associated pasture for five days. Milk production was measured on days 1 and 4 in the sole
cropped pasture and on days 1, 3 and 5 in the associated pasture. The production of
both groups was compared with that obtained by cows grazing pastures sown to B.
ruziziensis, the grass traditionally used on farms.
During the rainy season, pastures were rotated; the grazing cycle consisted of 4 or 5
days of occupation and 27 days of rest. During the dry season, grazing was continuous.
The Students (t) test (Steel and Torrie 1988) was used to compare the average milk
production. The averages of cows grazing the sole cropped pasture were compared with
the averages of cows kept on the native pasture and with those of cows grazing the associated pasture. The stocking rates of the native pasture, the sole cropped pasture and the
associated pasture were calculated using the method proposed by Paladines and Lascano
(1983), which takes into account animal units and grazing system.
Results
Forage availability and botanical composition
Table 1 presents mean DM availability and botanical composition of both pastures during the rainy season over three consecutive periods. The percentage of C. brasilianum
found in the associated pasture is so low that it was included in the percentage of
legumes present in the associated pasture.
Table 1. Forage availability and botanical composition of a pasture of Brachiaria brizantha cv. La Libertad alone and
associated with Arachis pintoi and Centrosema brasilianum during the rainy seasons of 199799 (Esparza, Costa
Rica).
Sole cropped pasture
Associated pasture
a
Grasses A. pintoi Legumes Weeds

(%)
(%)
(%)
(%)
Grasses A. pintoi Legumes Weeds

(%)
(%)
(%)
(%)
Year
T/ha
1997
4113
62
22
16
5105
57
18
13
12
1998
4483
67
18
15
5408
58
24
14
1999
6306
88
7155
54
42
Average 4967
72
16
11
5889
56
28
T/ha
a. Calopogonium mucunoides, Zornia spp. and Aeschynomene spp. were found in addition to C. brasilianum.
Forage availability was found to increase in pastures under recovery as the trial
advanced. At the beginning of the trial, forage availability in these pastures was 3 t/ha,
increasing to 6.3 t/ha in late 1999 in the sole cropped pasture with improved grasses and
to 7.1 t/ha in the associated pasture. This clearly shows the benefits of the method used
to recover these pastures.
50
Milk production of dual-purpose cows grazing B. brizantha alone and with A. pintoi
Increased forage availability is reflected in the increased stocking rate of pastures,

which in the case of native pastures was 1.5 AU/ha compared with 2.0 AU/ha for the
sole cropped pasture and 2.4 AU/ha for the associated pasture. Also, the percentage of
weeds was reduced in the associated pasture.
Milk production
The average milk production of cows grazing the native pasture and sole cropped pasture
of B. brizantha were similar (Table 2). However, the average production obtained on the
associated pasture was higher than that obtained on the native pasture (Table 3). Both
trials were carried out independently, using the native pasture as control.
Table 2. Average milk production of dual-purpose cows grazing native pastures and
Brachiaria brizantha during the rainy seasons of 199799 (Esparza, Costa Rica).
Milk production on two types of pasture
(kg/cow per day)1
Year
Native2
Brachiaria brizantha alone
1997
8.6
9.0
1998
7.5
7.7
1999
7.6
7.8
Average
7.9a*
8.2a
1. Average of 26 milking cows; 2. Mainly B. Ruziziensis.

* Averages in the same row followed by the same letter did not differ significantly
(P = 0.2271) using the Students (t) test.
Table 3. Average milk production of dual-purpose cows grazing native and associated pastures
during the rainy seasons of 199799 (Esparza, Costa Rica).
Milk production on two types of pasture
(kg/cow per day)1
Year
Native2
Brachiaria brizantha + Centrosema brasilianum
1997
8.6
9.3
1998
7.5
7.9
1999
7.6
8.1
Average*
7.9a
8.4b
1. Average of 26 milking cows; 2. Mainly B. Ruziziensis.

* Averages in the same row followed by the same letter did not differ significantly (P = 0.0246)
using the Students (t) test.
During 1997, the increase in milk production was higher compared with that of 1998
and 1999, because cows received a daily supplementation of 5 kg chicken manure and
1.5 kg of soybean hulls.
The average increase in milk production during the three years of the evaluation in
the sole cropped pasture as compared with the native grass was 3.8%, whereas the increase for cows grazing the associated pasture was 6.3%. Besides this increase in milk
production, producers income also increased because of the larger amount of milk sold.
51
Differences in milk production between sole cropped and associated pastures were
250 g/animal per day, which agrees with results found by Ullrich et al. (1994) who
worked with dual-purpose cows grazing B. decumbens alone and associated. At Escuela
Centroamericana de Ganadera (ECAG) in Atenas, Costa Rica, milk production of
Jersey dairy cows increased 9% when grazing B. decumbens + A. pintoi cv. El Porvenir pasture, as compared with cows grazing the sole cropped pasture and offered a commercial
concentrate as supplement (Romero and Gonzlez 1998). These results are similar to
those found in this study, despite these being dairy cows well known for their milk production that received concentrate. This confirms once more the beneficial effect of the
legumes on milk production.
Besides the benefits in milk production when improved pastures are used, there is a
significant increase in milk production/hectare as shown in Table 4.
Table 4. Average milk production (kg/ha) of dualpurpose cows grazing different types of pastures during
the rainy seasons of 199799.
Type of pasture
Milk production1
(kg/ha)
Native pasture2
11.6a
Sole cropped pasture
15.9b**
Associated pasture
19.7b***
1. Average production of 26 cows; 2. Native grasses,

mainly Brachiaria ruziziensis.
** p = 0.0001; *** p = 0.0253 (comparison of averages
using the Students (t) test).
Cipagauta et al. (1998) found that half-bred Holstein Zebu dairy cows, grazing B.
decumbens alone and associated with legumes (20% of the pasture) in the Amazon
piedmont of Caqueta, Colombia, produced 52% more milk/hectare on the sole cropped
pasture as compared with the native pasture, and 94% more milk on the associated pasture. The estimated milk production/hectare of cows grazing associated pastures differed
by 23% from that of cows grazing sole cropped pastures. Increases obtained for both pastures in the study are slightly lower (37% for the associated pasture; 70% for the sole
cropped pasture) as compared with the native pasture. The increase in milk production
between the associated pasture and the sole cropped pasture was 24%.
Gonzlez et al. (1996) found that Jersey cows, a local Central American race of dairy
cows, and crossbred cows between both of the former produced 14% more milk/hectare
in pastures of African star grass (Cynodon nlemfuensis) associated with A. pintoi than in
pastures of the grass fertilised with 100 kg N/hectare per year. This increase in milk production was lower than that found in our study, but the stocking rate was similar (2.6
AU/ha).
Lascano and Avila (1991) cite that the production of fat-corrected milk increased, on
average, by 20% in associations of B. dictyoneura with C. acutifolium and C. macrocarpum,
as compared with the grass pastures alone. In the case of Andropogon gayanus associated
with the same legumes, the increase in milk production of the associated pastures was,
on average, 15% as compared with the sole cropped pasture.
52
Milk production of dual-purpose cows grazing B. brizantha alone and with A. pintoi
with the same legumes, the increase in milk production of the associated pastures was,
on average, 15% as compared with the sole cropped pasture.
Conclusions
In this study the recovery of degraded pastures by establishing improved grasses increased milk production and pasture stocking rate. The introduction of legumes, such as
A. pintoi made it possible to increase milk production and pasture stocking rate more as
compared with native and improved sole cropped pastures. Pasture rehabilitation together
with the introduction of improved grasses and legumes, enhances forage availability and,
as a result, increases the pastures stocking rate.
References
Cipagauta M., Velsquez J. and Pulido J.I. 1998. Produccin de leche en tres pastures de
piedemonte amaznico del Caquet, Colombia. Pasturas Tropicales 20(3):29.
Gonzlez M.S., Van Heurck L.M., Romero F., Pezo D. and Argel P. 1996. Produccin de leche en
pasturas de estrella africana (Cynodon nlemfuensis) solo y asociado con Arachis pintoi o
Desmodium ovalifolium. Pasturas Tropicales 18(1):212.
Haydock K.P. and Shawn N.H. 1975. The comparative yield method for estimating dry matter
yield of pasture. Australian Journal of Animal Husbandry 15:663670.
Lascano C. and Avila P. 1991. Potencial de produccin de leche en pasturas solas y asociadas con
leguminosas adaptadas a suelos cidos. Pasturas Tropicales 13(3):210.
Paladines O. and Lascano C. 1983. Recomendaciones para evaluar germoplasma bajo pastoreo en
pequeos potreros. In: Paladines O. and Lascano C. (eds), Germoplasma forrajero bajo pastoreo
en pequeas parcelas. Metodologas de evaluacin. Red Internacional de Evaluacin de Pastos
Tropicales (RIEPT), CIAT (Centro Internacional de Agricultura Tropica), Cali, Colombia.
pp. 165183.
Romero F. and Gonzlez J. 1998. Produciendo ms leche mediante pastures asociadas con Arachis pintoi.
Hoja Informativa Tropileche 4 (marzo 1998). CIAT (Centro Internacional de Agricultura
Tropical), Cali, Colombia.
Steel R.D.G. and Torrie J.C. 1988. Bioestadstica: principios y procedimientos. 2nd ed.
Translated by Ricardo Martnez. McGraw-Hill, Mexico.
Ullrich C., Vera R. and Weniger J.A. 1994. Produccin de leche con vacas de doble propsito en
pasturas solas y asociadas con leguminosas. Pasturas Tropicales 16(3):2730.
53
Effect of regrowth age and cutting height

on productivity of Cratylia argentea
cv. Veraniega
M. Lobo and V. Acua
MAG, Costa Rica
Livestock development in the tropics, which is characterised as being extensive, is based

on the use of grasses; however, in farms that are a little more intensive, as is the case of
dual-purpose and semi-specialised dairies, animal nutrition is complemented with agroindustrial by-products during the dry season.
More and more, these by-products are not used because of their highly variable nutritive quality and price. Furthermore, in several of these regions these by-products must be
treated with heat, which affects their nutritive quality.
Therefore the use of shrub legumes has been studied in recent years to offer cattle a
better source of on-farm feed while reducing feeding costs.
Cratylia argentea cv. Veraniega is a shrub legume native to the Amazon region and
parts of Brazil, Peru and Bolivia. It is characterised by its adaptation to a broad range of
soils, mainly well-drained, low-fertility acid soils, and its capacity for regrowth during the
dry season. Its crude protein (CP) content is high, its in vitro dry matter digestibility
(IVDMD) is intermediate, and the consumption of this forage becomes more appealing
during the dry season.
This study aims to evaluate the productivity (production and nutritive value) of C.
argentea cv. Veraniega, at two ages of regrowth and three cutting heights, in Costa Ricas
central Pacific region.

At the beginning of the 1996 rainy season, a forage bank of C. argentea was established
on a farm in San Juan Grande de Esparza (250 metres above sea level (masl) with annual
average precipitation of 2400 mm and a six month dry season). Distance between plants
was 50 cm and distance between rows was 1 m. The soil was a clay loam Ultisol, with the
following characteristics: pH = 5.4, Ca = 7.2 cmol/litre, Mg = 0 2.6 cmol/litre, P = 6
mg/litre and OM = 3.1%.
One year after the bank was established and submitted to cutting during the dry
season, an area with 210 plants was selected for the present study. To determine the
productivity of cv. Veraniega, uniform cutting was performed at test heights in April
54
Effect of age of regrowth and cutting height on productivity of C. argentea
1998. Two ages of regrowth (60 and 90 days) and three cutting heights (30, 60 and 90
cm above the ground) were assessed.
At each cutting, the height of all plants was evaluated and 15 plants of each plot were
selected at random to perform the cutting at the corresponding age and height; the
number of sprouts in each and DM yield/plant was measured . The percentage of CP,
acid and neutral detergent fibre, and lignin contents were determined in the laboratory.
A Latin square design was used, and the averages were separated by the Waller-Duncan
test (Steel and Torrie 1988).
Frequent analyses were also made to determine the effect of the legume on soil fertility.

The effect of the age of regrowth and cutting height on yield of C. argentea during 1998
2000 is indicated in Table 1. All cutting heights tested tended to increase DM production per plant and per hectare. As cutting height increased from 30 to 90 cm, DM
production doubled. Likewise, the number of sprouts/plant tended to increase with
increasing cutting height.
Table 1. Average DM yield, plant height and number of sprouts of Cratylia argentea cv. Veraniega at two ages of
regrowth and at three cutting heights during 19982000 (Esparza, Costa Rica).
Age of
regrowth
(days)
60
90
Cutting height (cm)

Characteristics
30
60
Average
90
0.5
0.5
0.6
0.5
No. of
sprouts/plant
10.5
12.3
17.1
13.3
DM (g/plant)
89.7a* (40)3
110.6a (43.4)
178.1a (44.5)
126.1(42.6)
DM (t/ha)
1.3a (40.7)
1.7a (42.4)
2.7a (44.2)
1.9 (42.9)
Height of
regrowth
0.9
0.8
0.9
0.9
11.7
15.0
17.1
14.6
Height of
regrowth
No. of
sprouts/plant
DM (g/plant)
DM (t/ha)
256.9b (27)
3.9b (27.3)
304.9b (34)
4.6b (34.1)
494.9b (28.3)
352.2 (29.9)
7.4b (28.4)
5.3 (29.9)
1. Average of 11 cuttings; 2. Average of 8 cuttings; 3. In parentheses, percentage of total DM production during the dry
season.
* Averages of the same cutting height and rows followed by the same letters do not differ significantly (P = 0.05),
Average DM yields/plant at 60 days regrowth fell within the range cited by Argel and
Valerio (1996): 110190 g/plant in cuttings performed between weeks 8 and 14 at sites
with 5- to 6-month dry seasons. Yields obtained at 90 days regrowth averaged 352 g/
plant, being higher than those cited by the same researcher.
55
The average forage production at 90 days regrowth is over 4 t/ha more than that obtained at 60 days. In addition, average forage production during the dry season accounts
for approximately 43% and 30% of total production at the two ages of regrowth (60 and
90 days, respectively).
Figure 1 shows average DM production for the treatments. Within a single cutting
height, significant differences occurred between the two ages of regrowth, the yield
always being higher at 90 days regrowth. Table 2 shows the effect of age of regrowth and
cutting height on nutritive quality of C. argentea during 19982000. Forage quality at the
three cutting heights showed that the CP content tended to increase with increasing
cutting height from 30 to 90 days, and NDF and lignin contents. The CP content of
plants harvested at 30 cm height did not differ significantly at both ages of regrowth. On
the contrary, plants harvested at 60 and 90 cm height differed significantly (P = 0.08) in
CP because of the effect of age of regrowth.
DM (t/ha)
8
7
6
5
4
3
2
1
0
30
60
90
Cutting height (cm)

60 days regrowth
90 days regrowth
Figure 1. Average dry matter yield (t/ha) of Cratylia argentea at two ages of regrowth and three cutting heights during
19992000 (Esparza, Costa Rica).
The CP content in C. argentea varies broadly from 25% at 12 weeks of age (Franco
1997) to 20% at 8 weeks regrowth (Valerio 1994). Xavier and Carbalho (1995) reported
that, at 84 days, nitrogen content of C. argentea is low, approximately 3%, which is equivalent to 20% CP. These values are within the limits observed for other tropical legumes.
Franco (1997) found CP values of 23% at 2 months regrowth and values of 21% at 3
months regrowth. These values were higher than those found in the present study.
56
Effect of age of regrowth and cutting height on productivity of C. argentea
Table 2. Nutritive quality of Cratylia argentea at two ages of regrowth and three cutting heights during
19992000 (Esparza, Costa Rica).
Cutting height (cm)
Average
Age of regrowth
(days)
Quality
(%)
30
60
90
601
Crude protein
17.4a*
17.2a
17.7a
17.4
NDF
53.5a
55.9a
58.3a
55.9
ADF
43.1a
43.0a
44.1a
43.4
Lignin
14.8a
16.0a
16.1a
15.6
Crude protein
15.4a
15.0b
15.6b
15.3
NDF
58.6b
58.5a
61.2a
59.4
ADF
44.9a
44.7a
44.1a
44.6
Lignin
15.8a
16.3a
16.2a
16.1
90
1. Average of 11 cuttings; 2. Average of 8 cuttings.

* = Averages in the same column followed by the same letters do not differ significantly (P < 0.05),
Significant differences in NDF content occurred between the two ages of regrowth
when the cutting height was 30 cm (P = 0.0245). However, the highest NDF values
always occurred in cuttings at 90 days because the amount of woody material is greater
then (Figure 2).
Percentage
70
60
50
40
30
20
10
0
30
60
90
Cutting height (cm)

60 days regrowth
90 days regrowth
Figure 2. Average percentage of neutral detergent fibre in forage of Cratylia argentea harvested at two ages of regrowth
and three cutting heights (Esparza, Costa Rica).
57
In addition to the contribution of C. argentea to the quality of forage-on-offer, in the

analysis carried out in 1999 soil OM increased by 1% (3.1% in 1996 vs. 4.1% in 1999)
because of the high amounts of recycled material, especially leaves.
Conclusions
The following conclusions can be reached:
At 60 days regrowth, C. argentea produces good quality forage, regardless of the cutting height. Plant CP content was not affected by cutting height, but NDF content
was. At greater regrowth age, dry matter production was higher, protein content was
lower, and percentage NDF was greater.
C. argentea, cut at 90 days regrowth and at a height of 90 cm, produces good DM
production of acceptable quality.
Soil OM content at sites where C. argentea is established increased by 1% between
1996 and 1999.
Because of the high forage production of this legume during the rainy season, a
uniformity cut should be performed at the beginning of the rainy season to ensilage
material and use it during the dry season. In case this is not feasible, a uniformity cut
should be performed at the end of October to improve the quality and quantity of
forage during the dry season.
References
Argel P.J. and Valerio A. 1996. Cratylia argentea: un nuevo arbusto forrajero con potencial para el
trpico subhmedo. Paper presented at the 5th International Cycle on Tropical Pasture
Production and Research, Maracaibo, Venezuela.
Franco V. 1997. Evaluacin de la calidad nutricional de Cratylia argentea como suplemento en el
sistema de produccin doble propsito en el trpico subhmedo de Costa Rica. MSc thesis.
CATIE (Centro Agronmico Tropical de Investigacin y Enseanza), Turrialba, Costa Rica.
Steel R.D.G. and Torrie J.C. 1988. Bioestadstica: principios y procedimientos. 2nd edition.
Translated by Ricardo Martnez. McGraw-Hill, Mexico.
Valerio S. 1994. Contenido de taninos y digestibilidad in vitro de algunas forrajeras tropicales.
Agroforestera en las Amricas 1(3):1013.
Xavier D.F. and Carvalho M.M. 1995. Availacao agronmica de Cratylia argentea na zona de Mata
de Minas Gerais. Paper presented at the Cratylia workshop, 1920 July 1995, Brasilia, D.F.,
Brazil.
58
Effect of supplementing fresh and ensiled

Cratylia argentea cv. Veraniega on milk
production of dual-purpose cows
M. Lobo and V. Acua
MAG, Costa Rica
This study aims to evaluate the effect of the shrub legume Cratylia argentea cv. Veraniega,
on milk production when offered as fresh forage or ensiled to dual-purpose cows.

Trials were carried out on two farms located in Costa Ricas subhumid tropical Central
Pacific region, between 0 and 250 metres above sea level (masl), and well-defined dry
(DecemberApril) and rainy (MayNovember) seasons. The average temperature is 26C
and the average annual precipitation is 2500 mm.
Experiment 1
Six cows of third parturition and in second month of lactation were selected from a herd
of 45 milking cows that presented broad genetic variation, on a farm located in Barbudal
de Barranca, Miramar. Cows produced, on average, 5.5 kg milk/day.
After a 12-day period of adjustment to the new diets, each animal was offered the
forage supplement for 3 hours/day after milking. The trial lasted 30 days, during which
6 animals were rotated per treatment over a 10-day period, which was divided into 7 days
of adjustment and 3 days for individual measurement of milk production per cow. Milk
quality, especially in terms of fat and total solids contents, was determined.
The treatments evaluated were as follows: T1, control animals grazing the pasture
alone; T2, supplement of 12 kg sugarcane, 8 kg C. argentea, 0.6 kg semolina and 0.5 kg
molasses; and T3, 12 kg sugarcane, 3 kg chicken manure, 0.6 kg semolina and 0.5 kg
molasses.
The plants of C. argentea used had approximately 4 months regrowth, with an average
nutritive value of 16% CP and 33% DM.
A Latin square design was used in a changeover system (Lucas 1983), with three treatments and two animals per treatment. The separation of averages was performed by the
Waller-Duncan test (Steel and Torrie 1988).
59
Experiment 2
Six milking cows of third parturition were selected from a group of 11 animals with
similar breed characteristics (3/4 Swiss Brown and 1/4 Zebu) on a farm located in San
Miguel de Barranca, Esparza. All cows were in their second month of lactation and producing 5.5 kg milk/day on average. After milking, forage supplements were offered individually for 3 hours. The experiment design and animal management used were similar
to those of Experiment 1.
The treatments evaluated were as follows: T1, 12 kg sugarcane, 6 kg C. argentea silage
and 0.6 kg semolina; T2, 12 kg sugarcane, 6 kg fresh C. argentea and 0.6 kg semolina;
and T3, 12 kg sugarcane, 3 kg chicken manure and 0.6 kg semolina.
For T2, a heap silo was built for C. argentea plants at 4 months regrowth that had
been previously chopped. Molasses was gradually added at 10% weight basis of material
for ensilage as the silage was being compacted. Silage was first used 4 months after being
ensiled and its quality was determined by measuring pH, percentage of CP, DM content
and several organoleptic characteristics such as odour and colour.
Results
Experiment 1
Milk production differed significantly between treatments (P = 0.076) (Table 1). Higher
milk production was obtained when chicken manure was offered as protein supplement.
No significant differences occurred between treatments when chicken manure was
replaced with C. argentea. Milk production of cows not receiving protein supplement was
similar to that reached with T2. This could be attributed, among other factors, to the
high availability of fruits of the cohune palm (Acrocomia vinifera) existing in the pastures
and to which the cows had access, and the consumption of several other sources of feed
available during the dry season.
Table 1. Average milk production, total solids content, and cost:benefit ratio of dual-purpose cows receiving supplements
of different protein sources (Experiment 1).
Treatment
Milk production
Total
(kg/cow)
solids (%)
Fat
(%)
Cost of supplement
(US$/kg)
Price of milk
(US$/kg)
Cost:benefit
ratio
5.5b*
11.2
3.1
0.235
5.9a, b
11.5
3.2
0.109
0.241
2.2
6.3a
11.2
2.9
0.217
0.235
1.1
* Values in the same column followed by the same letter do not differ significantly (P = 0.076), according to
Waller-Duncans test.
The results of greatest interest in this study are reflected in the economic indicators,
for example, the cost of supplements, the income obtained by sale of milk, and the cost
benefit ratio. Replacing chicken manure with C. argentea reduces the costs/kg supplement from US$ 0.217 to US$ 0.11, thus doubling the benefits of the investment.
60
Effect of supplementing fresh and ensiled C. argentea on milk production
The consumption of C. argentea was good and was not rejected by the animals.
Experiment 2
Milk production was highest in the treatment in which fresh C. argentea was offered;
however, this treatment did not differ statistically from the treatment offering chicken
manure or from the treatment offering ensiled cratylia (Table 2).
Table 2. Average milk production, total solids content, and cost:benefit ratio in dual-purpose cows supplemented
with different sources of protein (Experiment 2).
Milk production
(kg/cow)
Total
solids
(%)
Fat
(%)
Cost of
supplement
(US$/kg)
Price of milk
(US$/kg)
Cost:benefit
ratio
5.1b*
12.3
3.6
0.164
0.260
1.6
5.5a
12.2
3.4
0.109
0.256
2.3
5.3a, b
11.7
3.0
0.217
0.245
1.1
Treatment
* Values in the same column followed by the same letter do not differ significantly (P = 0.08), according to
Waller-Duncans test.
The silage obtained was highly palatable and of good quality (pH = 4.5, CP = 16.5%
and DM = 36%). Silage quality can accordingly be ranked as excellent (Gonzlez et al.
1990). Milk quality did not differ among treatments; however, total solids and fat in
milk were slightly higher in T1 (ensiled C. argentea) and T2 (fresh C. argentea) than in T3,
which included chicken manure.
The lowest production costs occurred in T2 and the highest in T3. However, the
highest income per sale of milk was obtained with T1 and the lowest with T3. These
results indicate that the highest economic benefit was obtained with T2, followed by T1
and T3.
As in Experiment 1, the consumption of fresh and ensiled C. argentea was good and
the percentage of rejection by animals was 10%, mainly stems.
The experience with legume silage in the tropics is limited. In Cuba, Ojeda (1999)
found that grass/legume silage had good quality. Reagen (1999) found that Cassia silage
presented 12% CP and 58% IVDMD. In the process of ensiling legumes, fermentation
is mainly dominated by clostridia, which can be attributed to the high buffer capacity,
low soluble carbohydrate content in water, and, to a less extent, to the low DM content
(Pezo et al. 1989). Kass and Rodrguez (1987), cited by Pezo et al. (1989), observed a high
amount of volatile nitrogen (ammonium) and low levels of lactic acid in silage of
Gliricidia sepium.
Conclusions
Cratylia argentea, used as a protein supplement during the dry season, can substitute
chicken manure completely in milk cows, while it reduces the farmers dependency
on off-farm resources and improves the quality of the diet offered.
61
The nutritional quality of the milk produced by cows on diets of fresh or ensiled C.
argentea is better in terms of fat and total solids.

Profits were higher when milk production involved fresh and ensiled C. argentea than
when chicken manure was used because of the higher production and better-quality
milk.
Additional studies should be conducted on milk quality and maximum voluntary
consumption of cratylia. Different types of silos and silage systems should also be
studied.
References
Gonzlez D., Kass M., Ojeda F. and Wernli C. 1990. Conservacin de Forrajes. In: Ruiz M.E.
and Ruiz A. (eds), Nutricin de rumiantes, gua metodolgica de investigacin. IICA. Red de
Investigacin en Sistemas de Produccin Animal en Latinoamrica. Instituto Interamericano de
Cooperacin para la Agricultura (IICA), San Jos, Costa Rica. pp. 177237.
Lucas H.L. 1983. Design and analysis of feeding experiments with milking dairy cattle. 2nd Series
18. North Carolina State University, USA. Mimeograph.
Ojeda F. 1999. Harvesting and ensiling techniques. Estacin Experimental de Pastos y Forrajes
Indio Hatuey, Cuba. In: FAO Electronic Conference on Tropical Silage. 10 pp.
Pezo D., Kass M., Benavides J., Romero F. and Chves C. 1989. Potential of legume tree fodder
as animal feed in Central America. In: Devendra C. (ed), Shrubs and tree fodders for farm
animals. Proceedings of a workshop held in Denpasar, Indonesia, 2429 July 1989. CAB (Commonwealth Agricultural Bureau) International, Wallington, UK. 13 pp.
Reagen C. 1999. Wet season silage production at Taminmin High School. FAO Electronic Conference
on Tropical Silage. 4 pp.
Steel R.D. and Torrie J.C.1988. Bioestadstica: principios y procedimientos. 2nd ed. Translated
by Ricardo Martnez. McGraw-Hill, Mexico. 622 pp.
62
Effect of Centrosema macrocarpum

on milk production
D. Lara and K. Retegui
FUNDAAM and DEPAAM, Peru
The inclusion of forage legumes is a production strategy that allows production to be optimised and nutritional deficiencies of grass pastures to be overcome. This alternative
also makes better use of animal genotypes, land and producers capacities.
The Alto Mayo region covers 630,735 ha, of which 346,452 ha are suitable for grasslands. Of these 40% are not being used. A survey carried out in March of this year indicated that the Alto Mayo region has potential for livestock activities. Additional surveys
indicated that the region offers good possibilities of livestock production on pastures
planted to Brachiaria decumbens, B. brizantha and cut-and-carry grasses. However, Centrosema macrocarpum in protein banks and as cut-and-carry forage has proved to be a good
alternative to improve the diets offered to grazing animals.
The experiment aimed to:
1. measure the increase in milk production and the weight gain in calves attributed to
the use of C. macrocarpum to feed dairy cows; and
2. compare the farmers traditional system with a system based on Centrosema in terms
of economic merit.

Two 1-ha plots were planted on two farms located in Nueva Cajamarca and Soritor during September 1999. Planting density of Centrosema in both plots was high, with 0.4 m
0.4 m between plants and 8 kg seed/ha. A 95% coverage was reached at 5.5 months.
A completely randomised design was used with three treatments and seven cows/
treatment, over a 10-week period. The following model was used:
Yij = U + Tj + Eij
where:
Yij = milk production
U = overall mean
Tj = effect of the n nth treatment (i = 3)
Eij = experimental error
Treatments were as follows: T1, cut-and-carry grass (king grass) restricted to milking; T2, 50% amount of cut-and-carry grass provided in T1 and 50% of Centrosema
provided in T3, restricted to milking; and T3, 100% Centrosema restricted to milking.
63
Milking was done manually, once a day with calf on foot. After milking, calves were
allowed to suckle for 5 minutes and were then transferred to pens. Crossbred dairy
Gyr Holstein cows were used and left to graze on the pasture 6 hours/day, after
which they were taken to a pen where they received chopped king grass.
Control variables were: consumption of supplementary forage; milk production,
kg/cow per day during periods of maximum and minimum precipitation; production
costs of supplementary feed; and economic evaluation of results.
To measure consumption of supplementary forage (kg DM/cow per day),
Centrosema and king grass were separately cut, carried and chopped with machete and
supplied at milking. Forage on offer and forage rejected were monitored for the first
15 days of adjustment to determine the amount of forage needed during the experimental phase and thus estimate average intake per treatment.
Milk production (kg/cow per day) was measured four days per week. Based on this
report, production data had been determined for two periods of maximum and minimum precipitation.
Costs of producing supplementary feed considered establishment costs and maintenance costs per hectare and the cost of cutting and transporting forage.
The following evaluations were carried out to calculate total volume of forage (C.
macrocarpum and king grass)/hectare per year; forage availability (kg DM/hectare) in
random samples taken at the beginning of each grazing cycle; time of pasture recovery;
and pasture coverage at 20, 45 and 60 days after initiating grazing.
Economic merit was evaluated using the following formula:
EM =
Cost of feeding /cow period

Milk production (kg)
Results
Consumption of supplementary forage
Average consumption of DM at milking was 0.87, 0.96 and 0.90 kg/cow per day for T1,
T2 and T3, respectively, in addition to the voluntary consumption of Brachiaria during
grazing and king grass supplied in closed pens during the day and in the afternoon. A
trend to consume more grass + legume mixture was observed so it can be inferred that
the inclusion of the legume increases consumption by 23% in T2 and 11% in T3.
Milk production
The higher milk production was reached in cows supplemented exclusively with Centrosema (6.65 kg/cow per day), differing significantly (P < 0.05) with the treatment supplemented with Centrosema + king grass (5.56 kg/cow per day) and the treatment of only
64
Effect of Centrosema macrocarpum on milk production
king grass (4.11 kg/cow per day). Therefore supplementation with Centrosema had a
positive impact on milk production.
Effect of number of parturitions

No differences were observed in milk production between cows of two and three parturitions (4.66 kg/day) and cows of four and five parturitions (4.40 kg/day). Significant differences were observed, however, between these groups and first parturition cows (3.02
kg/day).
Effect of season on milk production

The average milk production of cows during the period of maximum precipitation was
5.13 kg/day, while it was 5.72 kg/day during the period of minimum precipitation, the
difference being significant (P < 0.001). This indicates the importance of the legume as a
nutritional supplement of cows during the dry season.
Economic merit
The cost/kg DM for each alternative was taken into account, which included the depreciation in a 6-year period of the establishment costs of Centrosema and king grass,
maintenance costs and harvest costs.
In the case of Centrosema, an establishment cost of US$ 380/ha was used, together
with an annual maintenance costs of US$ 40/ha, and a harvest cost equivalent to 0.6
personday per each 50 kg of fresh biomass. Annual productivity was estimated at 15 t in
5 cuttings. Thus, the total cost per kilogramme of DM was estimated at US$ 0.105.
In the case of king grass, estimated establishment cost was US$ 295/ha with annual
maintenance costs of US$ 110/ha and harvest costs equivalent to 1.4 persondays for
each 100 kg of fresh biomass. Annual productivity was estimated at 25 t in 4 cuttings.
Therefore, total cost per kilogramme of DM was about US$ 0.184.
The economic merit for each alternative appears in Table 1, which yielded indexes of
0.055 (T1), 0.056 (T2), and 0.047 (T3). Although production improved in T2 (32%) as
compared with T1, the economic merit of this treatment did not improve attributable to
the high consumption recorded in this treatment, which increased, in turn, the cost of
supplementary feeding. T3, however, presented a better economic index as compared
with T1 and T2.
65
Table 1. Calculating economic merit of several forage alternatives in the Alto Mayo Region
of the Peruvian Amazon.
Cost (US$/kg)
Feed consumed
(kg)
King grass
T1
0.87
T2
0.96
T3
0.90
Treatment
Centrosema
Total cost
0.16
0.160
0.10
0.098
0.196
0.178
0.178
Milk production/treatment
Economic merit
T1 Milk production = 4.11
0.037
0.035
0.027
Conclusions
Results indicate that supplementation with Centrosema had a positive impact on milk
yields when offered at milking to crossbred Gyr Holstein cows.
Supplementation with Centrosema also presented the best economic merit, ratifying
the important role this legume plays in improving milk production efficiency. The significant increase in milk production recovered additional expenses involved and maintained an additional margin of profit.
66
Effect of associating Brachiaria brizantha

with Arachis pintoi on milk yield
D. Lara and K. Reategui
FUNDAAM and DEPAAM, Peru
The inclusion of legumes in pastures of Brachiaria decumbens and B. brizantha in the Alto
Mayo Region of the Peruvian Amazon is a feasible alternative to increase animal productivity. This study therefore aimed to evaluate the effect of Arachis pintoi cv. Man
forrajero associated with Brachiaria on milk production and stocking rate of grazing
cows.

The treatments evaluated were: T1, B. brizantha in monoculture; and T2, A. pintoi associated with several species of Brachiaria.
Plots of the associated pasture, 4 ha in size, were established on a small farm in
Soritor. Initial availability of forage, milk production and stocking rate were recorded.
Crossbred Gyr Holstein cows were used in this evaluation. These same animals grazed
both types of pastures.
Each pasture was divided into two equivalent paddocks and animals grazed each paddock for 7 days. Milk production was measured in each paddock as of Day 3, during
maximum precipitation. Forage availability and botanical composition were measured
with random samplings in each paddock at the beginning of the grazing cycle.
Results
During the experimental period the producer controlled the rest/occupation periods
and the number of animals grazing. The stocking rate of each pasture was calculated on
the basis of these indicators: 2.07 AU/ha on pastures of B. brizantha and 4.13 AU/ha in
the associated pasture.
Forage availability and components of evaluated pastures indicated that total available DM in the associated pasture B. brizantha or A. pintoi was 7.59 t/ha at the beginning
of grazing, with legumes accounting for 6% and weeds for 11% of the botanical composition of the pasture. The pasture of B. brizantha alone produced 6.29 t DM/ha, with
9% weeds.
67
Milk production during the season of maximum precipitation was, on average, 5.98
kg/cow per day when the cows grazed the associated pasture, compared with 4.99 kg/
cow per day when they grazed the B. brizantha pasture.
Conclusions
Although these results are preliminary and correspond to a single evaluation, farmer expectations are high because the stocking rate of the associated pasture doubled that of
the sole cropped pasture and milk production per cow increased by 10%.
68
Research results
Activity 2.2 Evaluation of new feeding alternatives
to allow early weaning
Highlights
Supplementing pre-weaned calves with legumes during the dry season results in more
milk for sale and higher liveweight gain compared with calves grazing a sole cropped
grass pasture.
Use of Stylosanthes guianensis for strategic

supplementation of pre-weaned calves
J. Vela
IIAP, Peru
In the region of Pucallpa, in the Peruvian Amazon, milk price is very high due to a captive market of fresh milk. Pre-weaned calves usually consume between 1520% of milk
production from dam (residual milk). Therefore, developing a feeding alternative to partially substitute the milk consumed by pre-weaned calves with a high quality legume
could have a great impact on the income of smallholders because they could sell more
milk and improve their cash flow.

The following treatments were evaluated: T1, access of calves to native grass pastures for
8 hours/day, together with dam, and the remaining 16 hours in a pasture of Stylosanthes
guianensis + residual milk; T2, access of calves to native grass pastures 24 hours/day, without the presence of the dam + residual milk + concentrates; and T3, control (access to
native grass pasture 24 hours/day + residual milk).
Results
Table 1 presents trial results. Average daily weight gain of calves grazing stylo pastures
were similar to those obtained in the traditional pasture management system, but milk
sales increased by 23% (0.89 kg additional milk/cow per day). This was reflected in a
25% increase in income. The treatment with stylo + concentrate presented similar milk
sales to those of the treatment without concentrate, but higher daily weight gains. However, the variable cost was also higher, resulting in similar total net incomes as when only
stylo was used.
Conclusions
This improved feeding alternative can have significant impact on the quality of life of
small farmers. Stylo can be used to replace concentrates and thus increases cash flow and
income, and milk productivity without sacrificing a reduction in weight gain of preweaned calves.
71
Table 1. Average volume of milk sold/cow per day, liveweight increase of calves, and cash value of milk and beef per
treatment (Pucallpa, Peru).
Treatments
Item
Milk sold/cow per day (kg)
Milk price (US$/kg)
Daily weight increase (LW, g/calf)
Beef price (US$/kg)
Income from sale of beef (US$/calf per day)
Total value of production (US$/cowcalf per day)
Variable cost (US$/cowcalf per day)
Net income (US$/cowcalf per day)
72
Stylo
4.73
0.31
540
1.03
0.56
2.03
0.58
1.45
Stylo + concentrate
4.71
0.31
760
1.03
0.78
2.24
0.80
1.44
Control
3.84
0.31
530
1.03
0.55
1.74
0.58
1.16
Feeding pre-weaned calves

with Stylosanthes guianensis
in dual-purpose production systems
J. Velasquez, G. Ruiz and C. Lascano
CORPOICA and CIAT, Colombia
The main outputs of dual-purpose cattle production systems are milk and weaned calves
for fattening. Under traditional management, farmers usually prefer to sell as much milk
as possible to improve their cash flow, but, as a result of this practice, calf growth is slow
and mortality rates high. Therefore, the development of feeding systems that allow farmers to obtain more milk for sale and increase the number of pre-weaned calves with
adequate weight is a priority in this type of systems.
The use of Stylosanthes guianensis (stylo) for grazing pre-weaned calves has been tested
in Pucallpa, Peru (see information presented in this document). Results indicate that
with this alternative, farmers can sell almost one more litre of milk/cow per day and still
maintain adequate calf growth, which has important economic implications. This study
aimed to validate the results obtained in Pucallpa, using pre-weaned calves. The study
was conducted at the experiment station of the Colombian Corporation for Agricultural
Research (CORPOICA) in Macagual (Caqueta, Colombia).

Two groups of six calves each, aged 1 to 3 months, were used. One group of calves had
access to a 2-ha paddock of stylo after each milking and also received residual milk. The
six calves in the control treatment received milk equivalent to one-fourth of the udder at
milking and had access to a grass pasture after milking. In all cases calves remained with
their dam for 3 to 4 hours after milking, before accessing the grass or legume pasture,
depending on the treatment.
Results
The amount of milk for sale that resulted from the use of stylo for pre-weaned calves was
21% higher than that recorded with cows that had calves managed in the traditional
systems (Table 1). In addition, daily liveweight gain of calves with access to legume was
30% higher than that of the control group during the 90 days of the experiment.
73
Table 1. Volume of milk for sale and growth of pre-weaned calves with and without access to Stylosanthes
guianensis in Caquet, Colombian Amazons.
Treatment
Parameter
Milk for sale (kg/cow per day)
Liveweight gain of pre-weaned calves (g/day)
Control
3.3
297
Stylosanthes
guianensis
4.0
389
Conclusions
The results of this study agree with those obtained on small dairy farms of Pucallpa,
Peru. This technology could be very attractive to small dairy producers given that the
establishment cost of this legume is less than that of other alternatives based on legume
grass associations. Furthermore, cash flow increases because of the higher sale of milk,
without affecting calf liveweight gain.
In addition, this technology could form part of a croppasture rotation system to improve soil fertility, thus eliminating the need to fallow land.
74
Research results
Activity 2.3 Integration of information using ruminant
and feed system models
Highlights
Simulation models are useful tools to reduce research costs, predict nutritional
deficiencies and facilitate decision-making regarding research priorities in the area of
animal nutrition.
Decision tools to overcome nutritional

constraints in dual-purpose cattle in agrosilvopastoral systems: Net carbohydrate
and protein system model
F. Holmann and R.W. Blake
CIATILRI and University of Cornell
The Cornell Net Carbohydrate and Protein System (CNCPS) model was developed to
predict the nutrient requirements of cattle and feed utilisation by different types of
animals with diverse environmental and management conditions. It is a structure that
was designed to integrate and apply nutritional knowledge to:
solve feeding problems and reduce feed cost/production unit
set research priorities
design sensitive experiments
interpret experimental results
teach the application of biological principles through the integration of knowledge and
minimise nutrient excretion into the environment per unit of production.
The CNCPS simulates the effects of nutrient intake, ruminal fermentation, intestinal
digestion, absorption, and metabolism on nutrient utilisation and cattle performance.
Specific uses of the model are to:
predict the effects of feed composition and quantity on cattle performance
predict the effects of digestive and metabolic modifiers on cattle performance
evaluate and balance rations for the host animal and rumen bacteria
adjust cattle requirements and estimates of performance for different environmental
conditions and
simulate and predict the effects of gastrointestinal parameters on feed utilisation.
Thus, this model predicts nutrient requirements and nutrient pool sizes interacting
with changing diet composition available to animals.
Application of the model to research results at the

Quilichao Station
Data were collected from several trials with crossbred cows grazing Brachiaria decumbens
at three stocking rates (2, 3 and 4 AU/ha) and supplemented with sugarcane and varying
levels of Cratylia argentea.
77
Chemical composition included NDF, CP, lignin, IVDMD and DM for B. decumbens;
sugarcane and Cratylia for all trials. However, it was not possible to perform two important analyses that are also required by the model: solubility of protein and non-protein
nitrogen (NPN). Thus, without these values it was not possible to calculate the protein
and carbohydrate fractions. As an alternative, the values of solubility of protein and
NPN for B. decumbens generated by a PhD student at Cornell for the gulf coast of Mexico
were used to run the model and those for Gliricidia sepium (to emulate Cratylia) and for
sugarcane available in literature.
Other data from Avila (1999) included environmental and management variables.
Tables 1 to 3 include composition of diets, forage consumption and availability, and
actual milk yields vs. those predicted by the model.
Table 1. Chemical composition of Brachiaria decumbens, sugarcane and Cratylia argentea for the
different trials at Quilichao, Colombia.
Composition
Intermediate stocking rate
(3 AU/ha)
Dry matter (%)
Crude protein (%)
Neutral detergent fibre (%)
Lignin (%)
TDN (%)
Low and high stocking rate
(2 and 4 AU/ha)
Dry matter (%)
Crude protein (%)
Neutral detergent fibre (%)
Lignin (%)
TDN (%)
Sugarcane
Cratylia argentea
35.0
4.4
70.4
3.4
52.0
30.5
2.1
43.1
5.3
66.1
33.5
22.4
66.4
17.4
48.7
39.0
3.7
69.0
3.2
54.0
27.1
2.1
43.1
5.3
68.6
35.5
20.4
63.4
16.1
50.2
Table 2. Consumption of sugarcane and Cratylia and predicted intake of Brachiaria decumbens for the different
trials at Quilichao, Colombia.
Triala
Sugarcane
Cratylia
Brachiaria
decumbens
Total
DM
(% LW)
ISR + sugarcane
4.43
5.04
9.47
2.12
ISR + 75% sugarcane + 25% legume
2.64
1.16
5.81
9.61
2.15
1.48
2.28
5.83
9.59
2.15
0.45
4.16
5.03
9.64
2.16
LSR + sugarcane
2.00
7.48
9.48
2.30
LSR + sugarcane + legume
2.92
0.62
6.02
9.56
2.32
HSR + sugarcane
2.00
6.96
8.96
2.17
HSR + sugarcane + legume
2.92
0.70
5.60
9.22
2.24
a. ISR = intermediate stocking rate; LSR = low stocking rate; and HSR = high stocking rate.
78
Tropileche consortium
Decision tools to overcome nutritional constraints in dual-purpose cattle systems
Table 3. Actual and predicted milk yield using the CNCPS model for the different trials
at Quilichao, Colombia.
Actual milk yield
Predicted
milk yield
Energy
Protein
+ Sugarcane
6.0
4.9
1.4
+ 75% sugarcane + 25% legume
6.5
6.5
7.4
6.4
4.5
5.1

Low stocking rate (2 AU/ha)
6.6
1.9
3.1
7.9
7.3
7.9
8.2
8.1
8.1
+ Sugarcane
6.1
6.4
6.3
+ Sugarcane + legume
7.0
7.5
8.0
Trial
Intermediate stocking rate (3 AU/ha)
+ Sugarcane
+ Sugarcane + legume
High stocking rate (4 AU/ha)
The model predicts expected milk yield from three different sources: from metabolisable energy (ME), from protein (P) and from amino acids availability (AA). However,
the discussion will be centred on the results obtained from the first two sources because
there was no information on the AA composition of the forage obtained from Quilichao
nor was information available from the tropical forages library.
As can be observed, the CNCPS model predicted the observed milk yields in both
the low and high stocking rate trials, but failed to predict the observed milk yields for the
intermediate stocking rate trials. This under prediction came from both the energy and
protein portions, but the bias was larger in predicting milk yields using the protein
portion.
The probable cause of the low prediction of milk yields for the intermediate stocking
rate using the protein portion is the indigestible dry matter which causes low microbial
growth, due to high levels of NDF from both B. decumbens (6970%) and C. argentea
(6366%). Based on Dr Alice Pells comments, the high lignin content found in cratylia
could be the reason why the CNCPS model under-predicted observed milk yield. Based
on Pells experience, lignin content of cratylia leaves and stems should have been around
78% but not the 1617% as found in the Quilichao work. The secondary compounds
may be confounded with lignin, masking the overall effect of diet.
Based on the Quilichao results, several hypotheses can be drawn:
Cows were selecting B. decumbens with a higher CP content than actual lab results.
Digestion rates for carbohydrate and protein fractions used were wrong.
Information about the pool sizes is inadequate.
In addition, Table 4 presents the level of urea nitrogen found in milk (MUN) and the
level predicted by the model. The difference could be due to the soluble protein level in
the diet, which was higher than the CNCPS prediction.
79
Table 4. Level of urea nitrogen found in milk (MUN) and MUN levels predicted by
the CNCPS model for the various trials conducted in Quilichao, Colombia.
Trial
Intermediate stocking rate (3 AU/ha)
+ Sugarcane
Low stocking rate (2 AU/ha)
+ Sugarcane
+ Sugarcane+ legume
High stocking rate (4 AU/ha)
+ Sugarcane
+ Sugarcane+ legume
MUN observed
MUN predicted
10
11
14
17
3
3
7
11
8
12
0
0
6
12
0
1
Because of the slow digestion rates associated with the B3 protein fraction, the
CNCPS predicts that the neutral detergent insoluble protein (NDIP) makes little contribution to the rumen N pool. However, the NDIP contribution of tropical grasses to
rumen N balance may be higher than that predicted by the CNCPS model. There is a
probability that hypotheses (2) and (3) are valid.
Increasing the CP content in the diet does not increase the protein available for milk
yields in the same proportion because both the B. decumbens and the cratylia diets are
energy-deficient. However, the model was very sensible when the degradation rate of
fibre (B2 carbohydrate fraction) was changed.
To better understand why the model is not adequately predicting the results of intermediate stocking rate trials, we should have the following additional information: (1)
chemical analysis of feed refused (sugarcane + cratylia) during trials to better predict
nutrient pool sizes consumed; (2) body weights at weekly intervals for all trials and treatments to adjust for average tissue mobilised or deposited; (3) biomass availability of B.
decumbens after grazing (it was only measured at the beginning of grazing) to estimate
growth; (4) estimate of pasture growth by taking forage samples at two weeks intervals,
thus simulating cow grazing; (5) weekly milk yield and composition; and (6) body condition of animals.
Usefulness of the CNCPS model

The Tropileche Consortium can benefit from a close collaboration with the University
of Cornell by developing a broad database on tropical forages and by calibrating the
model for tropical conditions. Potential benefits of the CNCPS for Tropileche include:
This tool may help joint efforts to identify alternatives with potential to increase
animal productivity by identifying nutritional constraints, for example alternative
diets.
80
Decision tools to overcome nutritional constraints in dual-purpose cattle systems
It is the basis to evaluate on-farm production systems and propose alternatives to
select diets.
It compliments the CIAT farm-level model developed by R.D. Estrada to conduct ex
ante and ex post analysis at the farm and watershed levels.

It increases Tropileches menu of options to support and complement the activities of
partners and producer organisations, which could finance forage research.
High return on investment because Tropileche will invest marginally to complement
more than 10 years of research to develop the CNCPS at its current state.
81
Research results
Activity 2.4 Workshops and meetings to analyse and plan
research activities
Highlights
Workshops among Consortium partners facilitate co-ordination, increase the
efficiency of research, reduce risks associated with duplication of efforts, and facilitate
the establishment of research priorities for the short- and medium-term.
Annual meeting
February 1998, Costa Rica
F. Holmann, C. Lascano and A. Ramrez

CIAT, Colombia
In February 1998, the Tropileche Consortium held a workshop with all its partners in
Costa Rica to:
1. present the 1997 activity reports for benchmark sites (Costa Rica and Peru) and extrapolation sites (Nicaragua and Honduras)
2. define new activities for 1998
3. monitor research carried out in the field by Tropileche-Costa Rica and confront producers on the benefits of the new forage alternatives being evaluated
4. assess the need for strategic and participatory research as related to demand and
5. analyse new forms of collaboration with other national institutions in Costa Rica and
elsewhere in the region.
Twenty-one researchers from Peru, Colombia, Costa Rica, Nicaragua, Honduras and
the United States of America participated in the Tropileche Consortium workshop.
85
Annual workshop
June 1999, Peru
F. Holmann, C. Lascano, P. Kerridge and A. Ramirez

CIAT, Colombia.
The Tropileche Consortium held a workshop to plan and discuss its present and future
activities in South America from 27 June to 2nd July 1999, in Moyobamba, Peru. The
workshop aimed to:
1. present the research achievements obtained by the Consortium in Peru and pose
future research needs
2. define new activities to be conducted in Peru for 1999
3. participate in a field visit to understand and identify opportunities and needs of
animal production systems in the Alto Mayo region of the Peruvian Amazon
4. review strategic and participatory research based on needs and constraints and
5. analyse and discuss new forms of collaboration with other institutions in other
countries of South America, especially Ecuador, Bolivia and Brazil.
Invited participants included 22 researchers from Peru, Colombia, Ecuador, Bolivia
and Brazil.
86
Regional consultation meeting

October 1999, Costa Rica
F. Holmann, P. Argel, C. Lascano and A. Ramrez

CIAT, Colombia
A regional consultation meeting was held in Atenas, Costa Rica, to plan future activities
of Phase II of the Consortium in Central America and the Caribbean. The objectives of
the regional consultation were to:
1. present the status of Consortium activities and achievements to date
2. present and discuss opportunities to improve the dual-purpose production systems of
each of the six countries that will participate in Phase II of Tropileche
3. participate in a field visit to observe and discuss the technologies being promoted
among producers
4. develop profiles of research proposals for Central America and the Caribbean and for
each one of the six participating countries and
5. identify potential sources of funding.
A common vision of the future of regional production systems was developed at the
regional consultation meeting, and specific activities to be carried out by each country,
according to its specific priorities, were defined. There were 29 participants at the consultation meeting, including researchers, producers, and milk plant processing representatives from Costa Rica, Jamaica, Honduras, Nicaragua, Panama and Dominican
Republic.
87
Research results
Component 3
Usefulness of new forage systems
Activity 3.1 Diagnosis and economic characterisation

of land use at benchmark sites
Highlights
Ex ante analysis of improved technologies is useful to identify constraints to potential
adoption of new technologies by local producers.

Ex post analysis of new technologies is useful to determine the economic impact as a
result of the adoption process.
There is a high demand for new improved forage-based alternatives to intensify milk
production in the Alto Mayo region of the Peruvian Amazon.
The entry point to achieve greatest economic impact on small dual-purpose farms,
located in the dry subtropics, is through the establishment of shrub legumes with
sugarcane.
In Central America, the milk market for small artisan rural cheese factories has a
higher demand in the dry season than during the rainy season.
Small artisan rural cheese factories in Central America are willing to pay a higher
price for milk of better hygienic quality.
Ex ante analysis of new forage alternatives

for dual-purpose cattle farms in Peru, Costa
Rica and Nicaragua
F. Holmann
CIATILRI, Colombia
One of the objectives of the Tropileche Consortium is to develop new feeding alternatives
based on improved grasses and legumes. To achieve this, new germplasm generated at
CIAT is tested and validated with the participation of producers at different benchmark
sites. In addition, ex ante economic analyses are also performed to estimate the potential
impact of these new technologies and possible constraints to their adoption.
The objective of this study was to carry out an ex ante economic evaluation of new
forage alternatives available to producers in the lowland tropics of Latin America, using
as case studies the collaborating producers of the Tropileche Consortium in the forest
margins of Pucallpa (Peru) and the hillsides of the dry tropics in Esparza (Costa Rica)
and Esquipulas (Nicaragua).

Data for this study was obtained through direct interviews with each collaborating producer of the Tropileche Consortium at benchmark sites to understand their production
systems, use of resources, inputs and product prices, technologies used, and information
about the main characteristics of the watersheds where farms were located. For this
analysis, a CIAT-developed linear programming farm model was used. The model, based
on an electronic spreadsheet, maximises income.
Results
Table 1 shows several indicators for herds on dual-purpose cattle farms in Peru, Costa
Rica and Nicaragua. Larger herds were found in Costa Rica (47 cows and 72 AU), followed by Peru (31 cows and 50 AU) and Nicaragua (29 cows and 48 AU). Milk production was higher in Costa Rica (5.0 kg/cow per day) and Nicaragua (3.7 kg/cow per
day) and lower in Peru (3.0 kg/cow per day). These differences in milk production with
Peru can be attributed to (a) milk market restrictions and not to differences in animal
genotype, and (b) the low percentage of milking cows found in Peru (42%) compared
with Costa Rica (60%) and Nicaragua (58%).
91
Table 1. Averages of livestock inventory, milk production and land use in dual-purpose
cattle farms in Peru, Costa Rica and Nicaragua.
Variable
Livestock inventory (no.)
Milking cows
Dry cows
Heifers
Calves
Bulls
Total animal units (AU)1
Total milk production/cow (kg)
Milk production/cow (kg)
Milking cows (%)
Land use (ha)
Native pasture
Improved pasture
Agriculture
Forest/fallow
Total
Area under improved pastures (%)
Stocking rate (AU/ha)
Peru
(n = 9)
Costa Rica
(n = 7)
Nicaragua
(n = 4)
10.6
20.1
21.9
15.7
1.3
49.8
32.1
3.0
41.5
28.0
19.3
16.1
35.1
2.0
71.7
139.9
5.0
60.1
16.9
12.0
14.5
15.3
1.3
45.3
62.5
3.7
58.5
48.3
8.4
1.5
17.7
75.9
14.8
0.88
69.1
8.7
4.6
9.1
91.6
11.2
0.92
37.5
12.2
0.7
2.3
52.7
24.5
0.91
1. Cows = 1.0, heifers = 0.7, female calves = 0.3, bulls = 1.3.
Most of the farm area is in pastureland, ranging from 75% in Peru to 95% in Nicaragua. The highest forest area per farm ratio was found in Peru (23%) while in Nicaragua
this ratio was very low (4%). The area dedicated to agriculture in Peru and Nicaragua is
small (0.71.5 ha/farm) and limited to subsistence crops (rice, beans and maize); while
in Costa Rica this area is larger and more diversified (4.6 ha including rice, maize, beans,
sugarcane and fruit trees such as mango, cashew and melon).
Most of the area under pastures is covered with naturalised species, such as Hyparrhenia rufa in Costa Rica and Nicaragua; a small percentage is covered with improved
pastures (11% in Costa Rica, 15% in Peru and 24% in Nicaragua) in different states of
degradation. No pasture rehabilitation measures are applied, resulting in a low and
similar stocking rate for all countries: 0.9 AU/ha.
Resource prices and capital investment

Table 2 shows resource prices and capital invested in dual-purpose systems in each
country. The farmgate price of milk differs broadly among countries, ranging from US$
0.22/kg in Esquipulas, Nicaragua, to US$ 0.32/kg in Pucallpa, Peru. The price in Costa
Rica is intermediate between both the above prices: US$ 0.28/kg. The milk price received in Peru and Nicaragua is for raw milk while the price received in Costa Rica is for
milk cooled to 5C on the farm, which makes it a better quality product.
92
Table 2. Resource prices and capital invested in dual-purpose farms in Peru, Costa Rica and Nicaragua.
Peru
Variable
(n = 9)
Prices (US$)
Milk (kg)
0.32
Beef (kg cull cows)
0.60
Labour (day)a
4.40
Land (ha)
200
Pasture rental (per cow per month)
3.00
Cow (unit)
500
Heifer (unit)
450
Weaned calf (unit)
150
Bull (unit)
700
Establishment of improved grass (ha)
250
Establishment of Arachis with grass (ha)
340
Establishment of Stylosanthes (ha)
150
Establishment of Cratylia (ha)
400
Establishment of sugarcane (ha)
Invested capital (US$/farm)

Land
15,244
Livestock
29,561
Shed, corral and equipment
2000
Improved pasturesb
1050
Fencesc
4752
Total
53,147
Costa Rica
(n = 7)
Nicaragua
(n = 4)
0.28
0.60
8.80
2364
4.30
550
500
170
700
270
370
165
420
550
0.22
0.50
1.75
347
3.00
350
250
100
600
225
310
150
390
500
216,522
42,260
12,896
1175
6822
279,675
18,287
18,538
2125
1372
4597
44,919
a. Includes social benefits estimated at 24% for Peru, 43% for Costa Rica and 17% for Nicaragua.
b. Invested capital is estimated at 50% of establishment cost multiplied by the area under improved pastures on each
farm.
c. An average of 5280 lineal meters for the farms in Pucallpa, 7580 lineal meters for farms in Costa Rica, and 5108 lineal
meters for the farms in Nicaragua, with an average investment of US$ 0.90/lineal meter with four strands of wire
separated by posts spaced every 3 m.
The price of beef (as culled cows) was similar in Peru and Costa Rica(US$ 0.60/kg
LW), but lower in Nicaragua (US$ 0.50/kg LW). However, the price of the animals was
higher in Costa Rica, followed by Peru. Nicaragua had the lowest prices for beef and live
animals for all categories.
The labour cost also differs dramatically among countries, ranging from US$ 1.75/
day in Nicaragua to US$ 8.80/day in Costa Rica, including social benefits. These differences in labour cost affect the establishment cost of forage alternatives most, which
are higher in Costa Rica and lower in Nicaragua.
Similarly, the commercial value of land also contrasted, ranging from US$ 200/ha in
Pucallpa, Peru, to US$ 2364/ha in Esparza, Costa Rica. The main reason for this contrast lies in the high level of public infrastructure and proximity to markets found in
Costa Rica, and its long social and economic stability compared with Pucallpa (Peru) or
Esquipulas (Nicaragua).
93
All factors indicated above contribute to a higher capital investment per farm in
Costa Rica, with a commercial value of US$ 280 thousand/farm, compared with an
average value of US$ 53 thousand/farm in Peru and US$ 45 thousand/farm in Nicaragua. Of these figures, land and livestock are the main investment in all countries.
Therefore, the value of land accounts for 77% of the capital invested in Costa Rica, 44%
in Nicaragua and 28% in Peru, while livestock represents 15% in Costa Rica, 41% in
Nicaragua and 56% in Peru.
Production costs and income

Table 3 presents an estimate of direct production costs (variable + cash costs) during
1997, gross income, net cash flow and current profitability on invested capital. Labour
cost is the most important production cost in Peru (43%) and Costa Rica (63%), and the
second most important in Nicaragua (32%). This category includes family labour valued
as minimum wage. The second most important category is supplementation cost, except
in Nicaragua, where it was the most important cost.
Table 3. Direct production costs, gross income, net cash flow and family labour retribution in dual-purpose systems
in Peru, Costa Rica and Nicaragua.
Variable
Direct production cost (US$/farm per year)
Hired labour
Family labour
Total permanent labour (no./farm)
Animal supplementation
Animal health
Maintenance infrastructure and equipment
Others
Total
Cost/kg of milk
Cost of labour as % of total
Gross income (US$/farm per year)
Milk-derived income
Sale of culled cows
Sale of calves
Net cash flow
US$/farm per year
US$/ha per year
US$/cow per year
Retribution to family labour (US$/day)
Number of times minimum wage
Annual profitability on invested capital (%)
94
Peru
(n = 9)
Costa Rica
(n = 7)
Nicaragua
(n = 4)
257
1606
1.16
683
784
727
318
4375
0.29
43
6018
3643
970
1405
5586
3212
2.74
2848
224
1617
549
14,036
0.23
63
17,856
13,572
1490
2794
1155
630
2.83
2205
390
817
427
5624
0.20
32
6759
5019
760
980
1643
29
54
8.90
2.02
2.87
3820
49
81
19.27
2.19
1.37
1135
23
39
4.83
2.76
2.53
Total production cost of milk differed significantly in the three countries, ranging
from US$ 0.20/kg in Esquipulas (Nicaragua) to US$ 0.23/kg in Esparza (Costa Rica) to
US$ 0.29/kg in Pucallpa (Peru). The principal reason for the low production cost in
Nicaragua is because labour cost is 5 times lower than in Costa Rica and 2.5 times lower
than in Peru.
Production costs in Pucallpa, Peru, are high and above the international price of milk
(US$ 2000/t, equivalent to 130 g of powdered milk/litre of fluid milk, or US$ 0.26/kg
fluid milk), mainly because of the low milk yields/cow (3.0 kg/cow per day) and the
small proportion of milking cows (41.5%) regarding its counterparts in Costa Rica and
Nicaragua.
Regarding gross sales, most come from milk (60% in Peru, 76% in Costa Rica, and
74% in Nicaragua). The rest of income is represented by the sale of weaned calves and
culled cows.
Estimated monthly family income was US$ 270 for Pucallpa (Peru), US$ 586 for
Esparza (Costa Rica) and US$ 147 for Nicaragua. However, this income was approximately double the minimum wage in Peru and Costa Rica, but almost three times more
that in Nicaragua.
The profitability on invested capital during 1997 was very low in Costa Rica (1.37%),
followed by Nicaragua (2.53%) and Peru (2.87%). The reason why Costa Rica obtained
the least profitability despite having the highest family income and labour retribution
was because the farms in Costa Rica have a high commercial value (US$ 280 thousand/
farm) due to high land values, which is not the case for Peru or Nicaragua.
Ex ante analysis for Costa Rica

Figure 1 shows the production cost/kg milk for the different forage options. In Esparza
(Costa Rica), the production cost using traditional pasture (Hyparrhenia rufa) makes it
necessary for producers to supplement their milking herd during the 5-month dry
season. With an average production of 1350 kg/lactation, milk production costs were
US$ 0.31/kg, while the milk price received was US$ 0.28/kg. With the sale of weaned
calves, equilibrium was reached with a salary similar to the minimum wage.
Production costs decreased as productivity increased. Therefore, going from a lactation producing 1350 kg (5.0 kg/cow per day) to 1500 kg (5.55 kg/cow per day) reduced the costs/kg from US$ 0.31 to US$ 0.29. With a lactation producing 2000 kg
milk production costs decreased to US$ 0.23/kg, with cows grazing on H. rufa pastures
and receiving year-round supplementation of chicken manure and molasses to satisfy
those nutritional requirements that H. rufa cannot satisfy.
Cratylia with sugarcane in Costa Rica

With this forage option it is possible to completely eliminate the need to purchase feed
concentrates, molasses or chicken manure during the dry period. This option is capable
of maintaining the production even when lactating cows are producing 2000 kg/lac-
95
Feeding systems with legumes to intensify dairy farms
tation (7.4 kg/day). The production cost/kg of milk is reduced by 13% compared with
the current feeding system and 9% less with production levels of 1500 kg/lactation. The
investment required to implement this option in a farm with an average herd of 47 cows
in Esparza (Costa Rica) costs around US$ 6000, which includes the establishment of 8.9
ha of Cratylia, 1.8 ha of sugarcane and the purchase of a cane-chopper.
Cost (US$/kg)
0.35
0.3
0.25
0.2
0.15
1350
1450
1550
1650
1750
1850
1950
Milk/lactation (kg)
Jaragua + concentrate/chicken manure
Brachiaria + Cratylia + sugarcane
Jaragua + Cratylia + sugarcane
Brachiaria + Arachis + Cratylia + sugarcane
Figure 1. Milk production costs of different forage feeding alternatives in Costa Rica, assuming the same herd size.
Figure 2 shows the real interest rate that could be paid for this investment, depending
on the productivity per cow, assuming that the producer allocates 50% of the marginal
income obtained as a result of implementing this forage alternative. The real interest rate
in Costa Rica during the study was set at 13% (24% nominal interest minus 11% inflation rate) and the available credit was for a 5-year term with 1 year of grace. Under this
situation, it was not feasible to pay this credit unless cows achieved a production of 2000
kg/lactation. With productivities of 1500 kg/lactation, it would have been possible to
pay this credit if real interest rates had been lower (510%) and longer payback time
periods (close to 10 years).
This situation should be analysed for purposes of livestock policy and competitiveness. In an economic scenario of open markets operating without subsidies, producers
should have the option to obtain credits with real interest rates that reflect the oppor-
96
tunity cost of money at the international level, which is currently at 57% in real terms,
with 15 years to pay.
Real interest (%)
40
30
20
10
10
20
5
10
Payback period (years)

1350 kg/lactation
2000 kg/lactation
1500 kg/lactation
Figure 2. Real interest rate that could be paid by adopting Cratylia and sugarcane, based on
alternative milk yields in Costa Rica.
Other alternatives analysed, for example, establishing Brachiaria brizantha with or

without Arachis pintoi, yielded similar results. However, the establishment of these
options releases fragile lands that could be put to other alternative uses, such as reforestation. Figure 3 shows the percentage of pastureland that could be released for
other uses in Costa Rica.
Ex ante analysis for Nicaragua

The situation in Esquipulas, Nicaragua, was similar to Costa Rica; in other words, the
forage alternatives evaluated in this study significantly reduced production costs. Figure
4 shows the cost of producing milk with different forage options according to cow productivity. Under study conditions, milk production cost was US$ 0.26/kg while the
97
price received was US$ 0.22/kg. With the income obtained from the sale of weaned
male calves, producers obtain an income similar to the minimum wage.
Percent (%)
50
40
30
20
10
0
Cratylia +
sugarcane
Brachiaria + Cratylia
+ sugarcane
Brachiaria + Arachis + Cratylia +

sugarcane
Figure 3. Percentage of pasture area that could be allocated to alternative uses because of the adoption of
improved forage alternatives, with the same herd size in Costa Rica.
Cost (US$/kg)
0.3
0.25
0.2
0.15
0.1
0.05
0
1000
1500
2000
Milk/lactation (kg)
Jaragua + concentrate/chicken manure
Brachiaria + Cratylia + sugarcane
Jaragua + Cratylia + sugarcane
Brachiaria + Arachis + Cratylia + sugarcane
Figure 4. Milk production costs in Nicaragua, assuming the same herd size.
98
This situation, however, could improve even with the current situation if cow productivity was higher. The quality of H. rufa is enough to maintain cows yielding up to
1500 kg/lactation, without additional supplementation during the rainy season and
cows yielding 2000 kg/lactation, with energy and protein supplementation year-round,
but with H. rufa it is possible to reduce milk production costs to US$ 0.20/kg based on
grazing management (for example, good degree of coverage of pastures, weed control,
rotation of paddocks, shade, availability of water for grazing animals and adequate
resting period).
Cratylia with sugarcane in Nicaragua

With this forage option, and in the case of Costa Rica, it is possible to completely eliminate the need for supplementation during the dry season. The production cost was reduced by 31% (from US$ 0.26/kg to US$ 0.18/kg) with the same productivity of the
existing herd. Similarly, it is possible to reduce the production cost to US$ 0.14/kg with
a cow productivity of 1500 kg/lactation. It is even possible to reduce it to US$ 0.12/kg
with a cow productivity of 2000 kg/lactation. The investment required to establish this
forage option, on an average farm of 29 cows at this benchmark site, was approximately
US$ 4600. This investment covered the establishment of 5 ha of Cratylia, 2.4 ha of
sugarcane, and the purchase of a cane chopper with diesel or gasoline engine because
very little rural electrification infrastructure exists at this site.
Figure 5 shows the real interest rates that could be paid for this investment, depending on the productivity per cow, assuming that the producer allocates 50% of the marginal income to pay back the loan regarding the base scenario (H. rufa + feed concentrate). During the study, the Nicaraguan financial system offered a real interest rate of
18% for farm per livestock credits with a 5-year payback period.
In this situation and with the prevailing milk productivity at that time, it was not possible to adopt this forage alternative because it was not economically viable. However,
with productivities per cow of 1500 kg/lactation, the payment of a credit was perfectly
viable under the conditions prevailing in the country at the time of the study because it
was possible to pay a real interest rate up to 22%, with a payback period of 5 years. With
production levels of 2000 kg/lactation, the situation would have been even more viable.
The MAGWFP dairy development project in Nicaragua offers credits to small milk
producers at a real interest rate of 10% with a 5-year payback period, although the allowable maximum amount is up to US$ 3000/farm. This was an excellent opportunity
for small producers to adopt these new improved forage-based technological alternatives,
which allow them to significantly improve their competitiveness and income by reducing
production costs.
Ex ante analysis for Peru

The case of Pucallpa, Peru, is different from that of Costa Rica and Nicaragua. A characteristic of this region is the higher annual precipitation (2000 mm vs. 1200 mm in Costa
99
Rica and Nicaragua) and its even distribution. In consequence, the dry season in
Pucallpa is only 3 months long and not 6 months as in the case of Esparza (Costa Rica)
and Esquipulas (Nicaragua). As a result, the climatic conditions in Pucallpa allow
pastures to grow year-round.
Real interest (%)
60
50
40
30
20
10
0
5
10
1000 kg/lactation
1500 kg/lactation
2000 kg/lactation
Figure 5. Real interest rate which could be paid by adopting Cratylia + sugarcane, based on cow productivity in
Nicaragua.
Another important characteristic in Pucallpa is that the production per cow is very
low (3 kg/cow per day, Table 1), especially in the dry season when it is practically nil.
This characteristic can be due to several factors:
a) low genetic potential for milk production
b) little availability of nutrients in grasses due to high pasture degradation in low fertility
soils and
c) a very limited market for raw milk.
Regarding the last factor, daily production of fresh milk in the area of Pucallpa is
about 2500 kg and the cattle inventory decreased from 82 thousand heads in 1986 to 26
thousand at present as a consequence of terrorist activities and cattle rustling. The city of
100
Pucallpa, with a population of 300 thousand, does not have a milk processing plant. As a
result, consumption of milk in Pucallpa is mostly in the form of evaporated milk imported from other cities of the country. This situation has accustomed the population to
drink pasteurised milk. Thus, the market for raw milk is very small. Four of the nine producers who collaborate with the Tropileche Consortium thought that if they increased
their milk production, they would have problems to market the additional milk because
the market was already saturated. The other five producers sell their milk to the School
Milk Program, a public-funded project that provides fluid milk to children during school
hours.
Another factor that affects the potential to increase the milk market is the substitution of protein sources existing in the region. River products are a good example: a kg
of DM of fish protein costs US$ 2.90 in Pucallpa, while a kg of milk protein costs US$
8.80. In other words, milk protein is 303% more expensive than fish protein. Furthermore, there are other energy sources, such as cassava and rice, equivalent in dry matter
and significantly less expensive than milk. As a result, the market of raw milk in Pucallpa
is limited to the upper-class population and its possibilities for growth are very limited,
unless a milk processing plant can be established in Pucallpa to satisfy the demand for
evaporated milk that is now brought from Lima.
Base scenario versus alternative options in Peru

Figure 6 shows milk production costs under different forage alternatives. Unlike the
results from Costa Rica and Nicaragua, the most profitable option for Pucallpa under
the current situation in Peru is the base scenario found on farms. In other words, the
most competitive option is to maintain the herd in native pastures and provide milking
cows with a supplement of brewers yeast during the short 3-month dry season. Brewers
yeast is a viable option because it is abundant and inexpensive (US$ 0.15/kg DM, 22%
CP and 65% digestibility).
With this alternative, the milk production cost was US$ 0.33/kg while the price received was US$ 0.32/kg. Therefore, the additional income from the sale of male calves
increased farmers income a little above the minimum wage and paid for variable costs.
This alternative also proved to be the most attractive, with productivities of 1500 kg/
lactation and even up to 2000 kg/lactation. In no scenario did the forage alternatives
evaluated in this study (Brachiaria + Arachis or Brachiaria + Cratylia) succeed in reducing
milk production costs at levels below that of the native pastures supplemented with
brewers yeast.
Under the assumption that brewers yeast ceases to be a viable option for supplementation, another option evaluated was maize at US$ 0.23/kg. For the option of
Brachiaria with Arachis and/or Cratylia to form part of the solution, and with milk production levels of 800 kg/lactation, maize price would have to be US$ 0.38/kg (a 65%
increase) so that the cost of producing milk equals the current alternative. Therefore it
seems very unlikely that producers will adopt the alternative of Brachiaria with Arachis
and/or Cratylia if they have brewers yeast and/or maize at US$ 0.23/kg as alternatives.
101
With milking cows producing 1500 kg/lactation, the price of maize would have to
increase 9% to equal the production cost of both alternatives evaluated.
Cost (US$/kg)
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0
1500
800
2000
Milk (kg/lactation)
Native grass + brewers yeast
Brachiaria + Arachis
Figure 6. Milk production cost with different forage alternatives in Peru, assuming the same herd size.
The fundamental reason why none of the improved forage options were economically
better than the current management practice of native pastures + brewers yeast is
because of the high capital investment required per milking cow. The low current proportion of milking cows induces a high depreciation rate of improved pastures per cow.
In Pucallpa, the percentage of milking cows was 41% while in Costa Rica and Nicaragua
this figure was close to 60%. Therefore, to invest in these new improved forage alternatives, it would be necessary to increase the percentage of milking cows in a year to a minimum of 53%, or increase the stocking rate of 0.9 AU/ha to 1.3 AU/ha by introducing
more animals. In this new scenario, the evaluated forage options would form part of the
solution.
From the financial viewpoint, Figure 7 presents the real interest rates that could be
paid if a producer in Pucallpa invests in the establishment of B. decumbens associated
with A. pintoi. Peru had the highest real interest rate of the three countries considered in
this study, with 34% (44% nominal minus 10% annual inflation rate). Therefore, even if
the investment in these new forage options was economically superior, the high real interest rate available in Peru did not allow producers to adopt these technologies. Therefore, in the financial scenario existing during the study, producers in Pucallpa had no
option for intensification, (not even with productivities per cow of 2000 kg/lactation
and with payback periods of 10 years) because it was not possible to pay a real interest
rate of 34% (the best scenario possible was 15%).
However, the advantage of establishing these forage alternatives in Pucallpa is that
both options successfully released significant proportions of areas currently under
grazing so they could be destined to other alternative uses such as reforestation and/or
102
conservation (Figure 8). The area released represents 35.4% in the case of Brachiaria +
Cratylia (20.2 ha/farm), and 48.1% in the case of Brachiaria + Arachis (27.4 ha/farm).
Real interest (%)
20
15
10
5
0
5
10
15
5
10

1500 kg/lactation
2000 kg/lactation
Figure 7. Real interest rate that could be paid in Peru by adopting Brachiaria + Arachis based on alternative milk
yields.
Percent (%)
50
45
40
35
30
25
20
15
10
5
0
Brachiaria + Arachis
Figure 8. Percentage of pasture area that could be released for alternative uses in Pucallpa,
Peru, if improved forage alternatives are adopted, with same herd size.
103
Early adoption of Arachis pintoi

in the humid tropics:
The case of dual-purpose cattle systems in Caqueta,
Colombia
L. Rivas and F. Holmann
CIAT, Colombia
The region of Caquet, Colombia, is representative of the environmental and economic

situation prevailing in Colombias Amazon region. Development efforts began in this
region in the early 1900s and, have since gone through several phases up to the implementation of complex agricultural and cattle production systems that, over time, have been
progressively intensified. CIAT, together with Nestl, has assessed the evolution of livestock systems in Caquet regarding two topics of interest for both institutions: (1) the
adoption of forage technologies and (2) the production and productivity of livestock
systems.
The forage legume Arachis pintoi, or perennial forage peanut, is the result of a joint
research effort of CIAT with collaborating institutions. This material was evaluated during several years and released in 1992 as a forage alternative. The adoption and transfer
of new forage technologies are complex and time-consuming processes. The decision to
adopt an improved pasture involves a considerable investment of capital not only in
pasture establishment, but also in higher livestock requirements because this type of forage
technology increases the stocking rate. This is a critical issue for small- and medium-scale
farmers, such as those that exist in Caquet, with little capital and limited access to
financial institutions.
The present study focused on analysing the evolution of the early adoption process of
Arachis pintoi promoted by CIAT and Nestl to generate relevant information that would
allow possible limitations to adoption to be identified, while facilitating the design of
strategies that would accelerate the adoption process.

Basic data were collected through surveys in which the universe was the group of dairy
farms that supplied milk to the Nestl plant in Caquet. Two sampling strategies were
used:
a. A random sample whose size was estimated using conventional statistical methods
and
b. An additional sample of 52 farms that were already using the new material, in view of
the fact that the adoption of this technology was in an initial phase.
104
Early adoption of Arachis pintoi in the humid tropics
Sufficient information was gathered accordingly regarding experiences, difficulties

and prospects of producers testing Arachis.
Results
Available data shows that land use patterns in Caquet are highly dynamic when compared with the results of a study carried out in 1986 by Ramirez and Ser (Tables 1 and
2). The most relevant aspects of this comparative analysis were that (1) the relative importance of natural grasses of low productivity decreased as the use of improved pastures
increased; (2) the greater diversification of the forage germplasm used decreased the
relative importance of B. decumbens as compared with other types of Brachiaria; and (3)
unlike 1986, mixed pastures of grasses with legumes were observed as a new forage
alternative for Caquet.
Table 1. Land use dynamics in dual-purpose farms in Caquet, Colombia, during 1986 and 1997.
1986 survey
Land use
Total farm area
Total pasture area
Natural
Improved
Crop area
Fallow area
Forest area
Area (ha)
131
95
62
33
4
22
9
1997 survey
Area (ha)
100
73
47
26
3
17
7
158
129
38
91
3
10
16
%
100
82
24
58
2
6
10
Table 2. Change in area of improved pastures in Caquet (Colombia) during 1986 and 1997.
Species or cultivar
Brachiaria humidicola
Echinochloa polystachya
Brachiaria brizantha
Arachis in association
Arachis alone
Brachiaria dictyoneura
Axonopus scoparius
Other Brachiaria
Brachiaria ruziziensis
Hyparrhenia rufa (Puntero)
Panicum maximum (Guinea)
Brachiaria spp. (Par)
Pennisetum purpureum (elephant grass)
Other forages
Total
1986
1997
76.0
0.6
3.8
0.0
0.0
0.0
0.0
6.2
0.0
0.0
7.9
2.1
1.2
2.3
0.0
100.0
64.1
13.8
11.9
4.7
2.7
0.1
1.0
0.7
0.3
0.1
0.3
0.0
0.0
0.0
0.3
100.0
Variation (%)
11.9
13.3
8.1
4.7
2.7
0.1
1.0
5.5
0.3
0.1
7.6
2.1
1.2
2.3
0.3
0.0
105
Uses, knowledge and expectations about Arachis pintoi

Of 226 producers interviewed, most of them (179) had heard about the existence of this
new forage alternative in the region. Of this group, 171 had seen the material (76% of
all producers). At the time of the interview, 68 of them (30% of total) were using A.
pintoi. Two types of pastures with Arachis were identified: (1) seed banks, with an average
area of 1.3 ha/farm, and (2) pastures of grasses mixed with Arachis with an average area
of 9.6 ha/farm. Fifty-seven livestock owners (82%) who were experimenting with Arachis
used it mixed with grasses. Of a total of 68 early adopters, 21 (31%) had had problems
establishing peanut. The main problems were related to low germination rate of the grass
(8%), and slow establishment of the peanut (7%). However, the degree of satisfaction
with Arachis performance was high and 82% of farmers said they were satisfied with the
results obtained so far. The fact that more than half of the adopters (55%) have already
used the legume in associated pastures is significant.
Producers who have used Arachis have identified its capacity to increase the productivity of the associated grass as an important attribute of this new forage option.
About 63% of the producers have detected one or more of the following advantages:
increased stocking rate, higher milk yield and increased weight gain, enhanced competitiveness with weeds and improved palatability for animals. Table 3 summarises how
producers rate Arachis based on their experiences.
Table 3. Ranking of experiences with Arachis pintoi
by producers in Caquet, Colombia, in 1997.
Ranking
Excellent
Good
Regular
Bad
Do not know yet
Total
Producers
(no.)
10
19
4
1
4
38
Percent
(%)
26.3
50.0
10.5
0.6
1.5
100.0
About 12% of producers who have tested Arachis (8 of 68) are not satisfied with the
results obtained. Table 4 indicates the causes of this dissatisfaction.
Table 4. Causes of unsatisfaction from producers who have used
Arachis pintoi in Caqueta, Colombia, in 1997.
Producers
(no.)
Percent
(%)
Benefits not clearly identified
50.0
No progress observed
25.0
Not consumed by cattle
12.0
Cause of dissatisfaction
Tends to disappear
12.0
Total
100.0
106
Early adoption of Arachis pintoi in the humid tropics
Adoption rate during the study

The adoption rate of A. pintoi in Caquet, during the study period, expressed as the proportion of producers using this material, was close to 9.2% (16 early adopters out of a
random sample of 174 producers). Based on this adoption rate, a preliminary estimate of
total area planted to A. pintoi was about 3000 ha in the 2973 dairy farms supplying milk
to the Nestle plant. Of this area, 2626 ha (88%) were associations of A. pintoi with
grasses and the rest, seed banks of the legume.
Of 68 early adopters, 58 (85%) said they were willing to expand planted areas and
would increase, over the next three years, the area planted to Arachis by 11 ha/farm in
year 1, by another 10 ha/farm in year 2, and 11 ha more during year 3.
Technical and economic viability of A. pintoi-based

technology
The technology of grasslegume pastures, using A. pintoi, is economically attractive for
the Caquet region. Its profitability, expressed as internal rate of return, is substantially
higher than that of traditional technology based on improved grass pastures alone (Table
5).
Table 5. Profitability of Arachis pintoi mixed with different Brachiaria species in Caquet, Colombia, during 1996.
Type of pasture
Establishmet
cost
(US$/ha)
Milk yield
(kg/cow per
day)
Stocking
rate
(AU/ha)
Internal rate of
return
(%)
158
12.0
B. decumbens + A pintoi
282
3.5
1.5
19.3
B. humidicola + A. pintoi
337
3.5
2.0
21.8
368
3.5
2.0
21.1
B. decumbens
B. dictyoneura + A. pintoi
Changing traditional technology to a grasslegume association with Arachis substantially increases establishment cost from US$ 157/ha (B. decumbens alone) to US$ 368/ha
(B. dictyoneura + A. pintoi). In addition to the substantial increase in establishment cost,
the increase in stocking rate also implies the need to purchase more cattle.
In view of the current situation of Colombia, where high real interest rates prevail,
new and low-cost alternatives must be identified to establish improved forage alternatives.
Although the quality of commercial seed used in Caquet is high in terms of purity
and germination, its local price is high compared with neighbouring countries such as
Bolivia. The cost per kg of A. pintoi in the region ranged from US$ 20 to US$ 25/kg
during the study, while that of Bolivian seed cost around US$ 15/kg. Despite the attributes of A. pintoi, its cost is substantially higher than that of other forage legumes used
in the region, for example Pueraria phaseoloides (kudzu), Centrosema macrocarpum and
Desmodium ovalifolium, whose costs ranged between US$ 12/kg and US$ 15/kg.
107
The reduced economic activity in Colombia and the farmers scarce knowledge of
this new material affect its rate of adoption. Information on the use, management and
potential of Arachis should be diffused broadly to accelerate its adoption rate; seed costs
must also be lowered to reduce establishment costs.
108
The need for forage technologies in the Alto

Mayo Region of the Peruvian Amazon
S. Fujisaka,1 D. Lara,2 K. Reategui,1 J. Montenegro,3 R. Ventura,3 M. Diaz,4 R. Diaz4
and D. White1
1. CIAT, 2. FUNDAAM, 3. CTAR, 4. MAG-Peru
Researchers from the Fundacin del Alto Mayo (FUNDAAM), the Ministry of Agriculture (MAG) of Peru, the CTAR (Consejo Transitorio Agropecuario Regional), and
CIAT (International Centre for Tropical Agriculture) conducted a diagnostic field survey
to evaluate feeding systems for dual-purpose cattle in the Alto Mayo Region, Department
of San Martin, in the Peruvian Amazon.
The study was carried out in four communities (Nuevo Cajamarca, Soritor, Habana
and Rioja), where farmers owned, on average, some 2530 head of cattle, with 4 to 10
cows being milked at any given time throughout the year. Farm size ranged between 25
and 33 ha, of which 1618 ha were sown to pastures. The most common species was
Brachiaria decumbens, followed by B. brizantha. The milk produced in the region is mainly
marketed through a processing plant that sells milk to the government for a social
project providing milk to school children.
The invasion of weeds and wet-season trampling, which caused soil compaction, were
the most frequent pasture management problems. Parasites and diseases affected livestock. In addition to livestock production, the farmers also depend on subsistence crops
such as irrigated rice and coffee.
The survey indicated that the lack of improved forages was a constraint to enhanced
herd productivity in the region. In the four communities surveyed, several farmers had
established cut-and-carry forages, mainly elephant grass (Pennisetum purpureum) and king
grass (Pennisetum spp.). These farmers had also installed improved milking and feeding
stalls and infrastructure for animal feeding. Because of the low prices paid for raw milk,
many producers allocated the milk produced to late-weaned calves rather than to market.

The study area was located in the piedmont region of the Peruvian Amazon region, at
1000 metres above sea level (masl). Average annual precipitation was 1380 mm, with a
dry season between May and August. Mean annual temperature was 22C. The survey
covered 52 producers; interviews were informal, but structured around a series of themes.
Data were tabulated and analysed each day after field visits. Interview themes for the
following day were modified daily based on findings of the previous day. The research
109
team spent the last day of the intensive 4-day diagnostic survey process to draw conclusions and discuss implications for future needs of research and collaboration.
Results
Cattle numbers were determined on 18 farms in Nuevo Cajamarca and 22 farms in
Soritor. On the farms in Nuevo Cajamarca, the mean herd size was 32 animals, ranging
between 6 and 70. This number included 15 cows, of which nine were being milked
(Table 1). The ratio of cows to herd size and milking cows to dry cows were typical of the
study area.
Table 1. Livestock inventory in Nuevo Cajamarca and Soritor (Alto Mayo, Peru).
Nuevo Cajamarca
Cattle inventory
Soritor
Average
Range
Average
Range
Total
Categories
Cows
Milking cows
Bulls
32
670
26
565
15
9
1.4
260
030
05
11
4
1.3
327
017
03
Births
Deaths
Sales
11
1.5
4
030
04
0.14
8
2
5
221
010
017
Number of herd size varied. In Soritor and Nuevo Cajamarca, 30% of the producers
interviewed had up to 15 head of cattle (13 on average), 53% had between 16 and 49 (31
on average), and 18% had 50 or more head (62 on average). The ratio of small-, mediumand large-scale operations appeared to be characteristic of the whole area visited. Taking
Soritor (where herd size averaged 26 head) as example, producers reported, on average, 8
births (ranging between 2 and 21), 2 accidental or disease-related deaths (ranging between 0 and 10), and five sales during the year immediately before (ranging between 0
and 17) (Table 1).
Mean farm size ranged from 25 ha in Rioja and Soritor to 33 ha in Nuevo Cajamarca.
Farm size ranged from 4 to 90 ha, with herd sizes proportional to farm size. All farmers
had some land in pasture. Assuming an average area in pasture of 1618 ha among the
four communities, the mean proportions of pasture area in relation to the whole farm
varied from 56 to 64% (Table 2). Stocking rates ranged from a mean 1.3 head/hectare in
Habana to 1.8 head/hectare in Nuevo Cajamarca.
In Habana, 82% of the farmers had B. decumbens in their pastures, while 100% of the
farmers in Rioja and Soritor had pastures sown to this grass, accounting for 7293% of
total pasture area. In Rioja, 18% of the farmers had also sown B. brizantha and 42% in
Soritor (Table 2).
110
The need for forage technologies in the Alto Mayo Region of the Peruvian Amazon
Table 2. Land use at study sites in Alto Mayo, Peru.

Nuevo
Cajamarca
(n = 18)
Soritor
(n = 22)
Land use
Area
(ha)
Area
(ha)
Farm size
Pasture
B. decumbens
B. brizantha
Cut & carry
King grass
Elephant grass
Coffee
Lowland rice
Forest/fallow
Stocking rate
Chopper
33
18
13
1
3
1
2
5
2
2
1.8
100
100
89
33
100
66
78
55
28
55
22
25
16
16
2
0.4
0
0
2.4
3
6
1.6
50
Habanan
(n = 11)
%
Area
(ha)
100
100
100
42
40
72
0
55
21
75
32
18
14
3
0.6
0.5
0.14
2
0.6
11
1.3
Rioja
(n = 11)
Area
(ha)
100
100
82
27
64
55
18
73
36
100
45
25
16
15
0.5
0
0
0
1
2
3
1.7
100
100
100
18
27
27
0
73
73
73
Many farmers had also established cut-and-carry forages: 27% of farmers interviewed
in Rioja and 100% of the farmers in Nuevo Cajamarca. Areas sown ranged from small
plots in Rioja, Soritor, and Habana (0.5 ha on average) to 3 ha in Nuevo Cajamarca
(Table 2). Varieties sown for cut-and-carry feeding were mainly king grass and elephant
grass. None of the farmers in Rioja and half of those in Soritor had mechanised forage
cutters.
Other main land uses were coffee, irrigated rice, annual crops, and forest or forest
fallow. Between 50 and 66% of farmers in each community grew coffee on plots of 1 ha
(in Rioja) to 5 ha (in Nuevo Cajamarca). From 21 to 73% of the farmers grew lowland
rice on plots from 0.6 to 3 ha in size. More than half the farmers had land under forest
or fallow. Habana was the most forested (11 ha/farm) (Table 2).
Animal health problems were ectoparasites such as ticks and nuche, (a skin burrowing maggot) and diseases such as mastitis and stomatitis. Treatments consisted in
farmers administering prophylactic drugs.
Weed control was the problem associated with pasture management that demanded
most labour. Farmers in flatter areas suffered from soil compaction caused by pasture
trampling in the wet season. Although paddocks were regularly rotated, some farmers
rented additional pastureland during the rainy season to decrease pasture degradation.
Additional pastureland was seldom leased during the dry season (Table 3).
Farmers interviewed were consistent in terms of what they wanted to achieve in the
future. Around half wanted to increase their areas sown to pastures, and an equal number of farmers wanted to increase and/or improve their herds. Only a few mentioned
plans to decrease herd size, but planned to improve herd quality by introducing improved breeds and selection. In Nuevo Cajamarca, 50% of farmers wanted to improve
their feeding/milking sheds. From 3366% of farmers wanted to improve their pastures
by establishing B. brizantha and other new forage species, including legumes (Table 4).
111
Table 3. Frequent problems1 in the Alto Maya Region of Peru and percentage of farmers
identifying these problems.
Site
Problem
Livestock pests and diseases
Weeds
Compaction of pastures
Crop pests and diseases
Lack of land
Non-agricultural aspects/others
Nuevo
Cajamarca
94
44
22
56
23
78
Soritor
Habana
Rioja
95
55
18
23
9
33
640
18
55
36
18
27
100
18
9
45
36
63
1. Low prices paid for products, land tenure and poor water quality.
Table 4. Farm improvement plans in the Alto Mayo Region of Peru and percentage of farmers
interviewed who want changes.
Site
Improvement plans
Increase the area under pasture
Improve sheds/corrals
Improve the herd
Improve the pasture
Other
Nuevo
Cajamarca
Soritor
Habana
Rioja
50
55
55
66
50
40
27
59
59
22
45
18
45
36
36
64
36
64
55
45
Conclusions
The main conclusions of the activities carried out by this collaborative effort between
national institutions and CIAT in the Alto Mayo Region of Peru were as follows:
Local farmers managed a mixture of enterprises, changing proportions over as time
determined by market variables.
Farmers adjusted the amounts of milk sold versus the milk consumed by calves in
response to prices paid for milk.
Farmers had technically sound but expensive solutions to their problems of weeds
and cattle diseases.
A high number of farmers were in process of adopting cut-and-carry forages, improving feeding and milking sheds and installing forage cutters. The demand for
improved forages was therefore relatively high.
Although the need for dietary protein was acknowledged, the adoption of forage
legumes has been hindered by high losses in biomass due to attacks of insects and
wild herbivores.
Very little pasture degradation due to pasture trampling by animals was observed at
the time of the survey, although producers mentioned that this was a production
constraint in flatter areas.
112
The need for forage technologies in the Alto Mayo Region of the Peruvian Amazon
Recommendations
Based on results and conclusions, the group recommended additional research on basic
indicators, weeds, cattle pests, improved forages and soil macro-fauna.
Recommendations included the following:
Basic indicators must be defined and monitored to better understand local dualpurpose systems and to measure future changes. The group proposed the monitoring
of a sample group of dual-purpose cattle farms to record data on herd composition,
birth and death rates, milk production, liveweight gain and feeding systems.
Research on weed control in pastures should be strengthened, especially regarding
costs. Further research is needed on the effects of periodic burning, pasture rotations,
stocking rates, and pasture species on weed composition and spread. Pasture species
that help control weeds and that can be used in livestock or pastures associated management systems should be evaluated.
Farmers continually invested in treatments to control parasites in animals. Factors
favouring the appearance of these pests should be studied.
FUNDAAM and the Ministry of Agriculture have conducted substantial research on
forage improvement. The Ministry recently distributed seed of C. macrocarpum, P.
maximum and B. brizantha to some 200 farmers for testing and multiplication. The
goal was that each farmer would be able to establish a protein bank. Forage species
and varieties that not only improve animal and milk outputs, but also help suppress
weeds, are resistant to drought, and are tolerant to water logging should also be
identified. Farmers recognised the potential value of forage legumes, although they
are attacked by pests and rodents. The group suggested enhanced farmer participation in the testing of a cocktail of forage varieties and species.
The group agreed that there is need to better understand political decisions and the
market context relevant to dual-purpose livestock producers. The issues discussed included national research priorities and policies, building of roads, the Glass of Milk
programme for local school students, credit, and national policies concerning milk
imports, exports and subsidies.
113
Potential benefits of new forage germplasm

in dual-purpose cattle systems in the dry
tropics
F. Holmann
CIATILRI, Colombia
Milk production in Central America and the Caribbean has grown during the 1990s at
an annual rate of 4.6%, which evidences the marked dynamism of the sector (Umaa
1998). However, even with this high growth, the region is a net importer of dairy products. These imports are growing at an annual rate of 13% due to the high demand. The
region went from importing 20% of its needs in 1990 to almost 28% in 1996, equivalent
to US$ 104 million/year (Umaa 1998).
It is estimated that around 85% of the milk produced in the region comes from dualpurpose farms that produce milk and beef. This type of production system is mostly
found in small and medium farms where pastures are the main feeding source because
they are inexpensive and abundant.
Research executed by CIAT and ILRI carried out by the Tropileche Consortium in
the subhumid areas of Costa Rica has demonstrated that improved forages contribute to
the intensification of milk production on small dairy farms by: (1) increasing milk production and stocking rate; (2) releasing areas not suitable for livestock for use as environmental reserves; and (3) reducing the demand for purchased feed inputs and thus improving the cash flow (Holmann et al. 1999). In addition, producers outside benchmark
sites have been observed to spontaneously adopt new forages, mainly in Costa Rica and
Peru and to a lesser extent in Nicaragua and Honduras.
Objectives
This study aims to measure the impact of new forage germplasm on productivity and
reduction of production costs, at different on-farm levels of adoption; identify the potential benefits; and determine the regions seed production needs. The study covered
dual-purpose farms located in hillside areas where the Tropileche Consortium operates
in the dry tropics of Costa Rica, Honduras and Nicaragua.

Data for this study were obtained from direct interviews with 78 producers located in
the dry tropics of each country to understand their systems of production, their use of
resources, the prices of inputs and products, and the technologies used. In Costa Rica,
114
TropilecheConsortium
Potential benefits of new forage germplasm in dual-purpose cattle systems in the dry tropics
21 producers were randomly interviewed in the area of Puntarenas and Esparza in the
Central Pacific region. In Honduras, 20 producers were interviewed in the areas of
Olancho, Catacamas and Juticalpa. In Nicaragua, 37 producers were interviewed in the
Muy-Muy and Esquipulas watersheds (Department of Matagalpa), where livestock production is an important activity.
A CIAT-developed linear programming simulation model was used to analyse information. This model, based on electronic sheets, maximises on-farm income.
Productivity and land use

Table 1 shows the averages for livestock inventory, milk production, reproductive parameters and land use in dual-purpose farms in Costa Rica, Honduras and Nicaragua during the time of the study.
The largest mean herd size was found in Nicaragua (44 cows and 81 AU) followed by
Honduras (42 cows and 78 AU); the smallest mean herd size was found in Costa Rica
(23 cows and 36 AU). Milk yield, however, was highest in Costa Rica (6.8 kg/cow per
day during the rainy season), followed by Honduras (6.6 kg/cow per day) and Nicaragua
(4.2 kg/cow per day).
Reproductive and animal management parameters are similar in the three countries
with the only difference that calf and adult mortality is highest in Nicaragua.
On-farm land use differed among countries. Large parts of the farms were allocated to
pastures, ranging from 77% in Honduras to 95% in Nicaragua. The highest proportion
of forest was found in Costa Rica (13%) while this area was minimal in Nicaragua (3%).
Small areas were allocated to agricultural crops in Costa Rica and Nicaragua (0.61.8
ha), being much larger in Honduras (7.6 ha).
The highest percentage of pasture area under low-productivity naturalised species
(96%) was found in Nicaragua, followed by Costa Rica (84%). The pasture area in
Honduras had the highest proportion of improved varieties (69%), which explains to a
great extent the high stocking rate found in Honduras (1.44 AU/ha) compared with the
other two countries (0.75 AU/ha for Costa Rica and 0.99 AU/ha for Nicaragua).
Prices of inputs/outputs and use of feeding

supplements
Table 2 contains the prices of inputs and outputs used by dual-purpose farms in each
country. The farmgate price of milk paid to producers is different in each country and
varies depending on the season of the year. The average weighed price is higher, however, in Honduras (US$ 0.26/kg), followed by Costa Rica (US$ 0.22/kg) and Nicaragua
(US$ 0.21/kg). This difference in price could be attributed to the fact hat Honduras has
a milk deficit, while Costa Rica and Nicaragua export milkCosta Rica in the form of
long-life UHT milk and Nicaragua in the form of fresh, non-pasteurised cheeses.
115
Table 1. Averages of livestock inventory, milk production, reproductive parameters and land
use in dual-purpose cattle systems in Costa Rica, Honduras and Nicaragua.
Costa Rica
Variable
(n = 21)
Rainy season
Milking cows (no.)
16
Dry cows (no.)
7
Milk production/farm (kg/day)
109
Milk production/cow (kg/day)
6.81
Dry season
Milking cows (no.)
16
Dry cows (no.)
7
Milk production/farm (kg/day)
66
Milk production/cow (kg/day)
4.12
Livestock inventory (no. of animals)
Heifers
6
Calves
13
Bulls
1
1
Total AU
36.4
Reproductive parameters
Calving rate (%)
71.4
Culling rate (%)
17.9
Adult mortality (%)
3.7
Calf mortality (%)
6.3
Duration of lactation (days)
285
Land use (ha)
Native pastures
40.5
Improved pastures
7.7
Agriculture
0.6
Forest/rest
7.2
Total
56
0.75
1. Cows = 1.0, heifers = 0.7, calves = 0.3, bulls = 1.3.
Honduras
(n = 20)
Nicaragua
(n = 37)
30
12
198
6.6
33
11
139
4.22
28
14
136
4.86
23
21
77
3.35
36
29
1.8
78.2
36
22
4
81
71.4
14.8
1.7
6.7
286
75
18.2
4.8
13.6
274
13.7
30.1
7.6
5.6
57
1.44
78.5
3.2
1.8
2.5
86
0.99
During the rainy season the milk price varies from US$ 0.27/kg, paid by processing
plants in Costa Rica, to US$ 0.18/kg in Honduras. During the dry season, intermediaries in Honduras pay the best milk price (US$ 0.34/kg), whereas intermediaries in
Nicaragua pay the lowest (US$ 0.17/kg). The largest buyers of milk were processing
plants in Honduras (65%) and Costa Rica (42%), while in Nicaragua intermediaries
purchased 46% of the milk produced.
The price of beef (as culled cow) was similar in Costa Rica (US$ 0.62/kg), and
Nicaragua (US$ 0.65/kg), while it was higher in Honduras (US$ 0.80/kg), where the
commercial price of animals is significantly higher.
116
Table 2. Prices of inputs and products on dual-purpose cattle farms in Costa Rica, Honduras and
Nicaragua.
Variable
Milk
Weighted price (US$/kg)
Rainy season
Processing plant
Intermediaries
Farmgate
Dry season
Processing plant
Intermediaries
Farmgate
Producers (%) sell milk to
Processing plant
Intermediaries
Other producers
Meat (US$/kg culled cow)
Use of labour
Permanent (no. of people/farm)
Temporary (no. of persondays/farm
per year)
Labour cost (US$/day)1
Efficiency (no. of cows/person)
Land value (US$/ha)
Animals (US$/animal)
Cow
Heifer
Weaned calf
Bull
Input prices (US$/kg)
Chicken manure
Molasses
Commercial feed concentrate
Salt
Minerals
Costa Rica
(n = 21)
Honduras
(n = 20)
Nicaragua
(n = 37)
0.22
0.26
0.21
0.27
0.17
0.20
0.18
0.21
na
0.21
0.17
0.18
0.27
0.17
0.21
0.31
0.34
na
0.24
0.27
0.28
42
29
29
0.62
65
35
0
0.80
11
46
38
0.65
1.52
4
2.95
31
3.36
27
9.3
17.3
3676
3.6
15.2
2078
540
500
160
600
750
415
140
800
0.04
0.08
0.19
0.13
0.94
na
0.09
0.21
0.12
0.86
2.3
13.2
612
485
330
110
550
0.065
0.08
0.19
0.12
0.95
1. Includes social benefits estimated at 44% for Costa Rica and 17% for Honduras and Nicaragua.
na = not available.
The use and cost of labour also differs markedly among countries, ranging from US$
2.30/personday in Nicaragua to US$ 9.30/personday in Costa Rica (including all social
benefits). These differences in labour cost affect efficiency, this effect being greatest in
Costa Rica (17.3 cows/person) with the smallest herd size and lowest in Nicaragua (13.2
cows/person) with larger herd sizes.
117
Similarly, commercial land value also contrasts, ranging from US$ 612/ha in Nicaragua to US$ 3676/ha in Costa Rica. The high land value in Costa Rica can be attributed to the high level of public infrastructure, the proximity to urban markets, and the
countrys long social and economic stability.
The amount of feed supplements used/cow per year in Costa Rica is greater than
those ones reported by the other countries (Table 3), which is consistent with the higher
milk yields and efficiency of labour reported in Costa Rica.
Table 3. Amount of feed supplements used on dual-purpose cattle farms in Costa Rica,
Honduras and Nicaragua.
Consumption (kg/cow per year)
Feed supplement
Chicken manure
Commercial feed concentrate
Rice polishing
Molasses
Salt
Minerals
Costa Rica
(n = 21)
Honduras
(n = 20)
Nicaragua
(n = 37)
496.7
508.6
137.3
41.3
38.1
19.0
na
292.0
na
87.3
7.8
20.9
410.6
na
na
70.5
24.1
12.4
na = not available.
Analysis of potential benefits of new forage

germplasm
This study aims to estimate the potential benefits of new forage germplasm, with different levels of adoption on-farm and at the regional level in the dry tropical hillside
areas where the Tropileche Consortium operates. The alternatives analysed were
Brachiaria brizantha, B. decumbens, and B. dictyoneura, and the legumes Cratylia argentea
and Arachis pintoi, and sugarcane as energy source.
Table 4 presents the nutritional parameters and biomass production of all forage
alternatives evaluated in this study using the simulation model. Animal management
parameters were taken from farm averages in each country so that they represented the
current conditions of management and resource use (Tables 1 to 3).
The situation prevailing during the study was evaluated for each region within each
country and this, in turn, was compared against three levels of adoption. Similarly, the
following parameters were estimated to compare the different scenarios:
additional investment in the new germplasm (per farm and per cow in the herd)
net income (per hectare and per cow) after discounting fixed and variable costs and
use of family labour valued as minimum wage cost

milk productivity/hectare
livestock area that can be released as a result of the different levels of adoption (in
number of hectares and as percentage of farm area allocated to livestock production)
118
production cost/kg milk as the maximum expression of competitiveness.

Table 4. Forage parameters used to run the simulation model at different benchmark sites of the Tropileche
Consortium.
Forages
Parameters
Crop duration (years)
Rainy season
Edible biomass production
(DM, t/ha)
Crude protein, CP (%)
CP degradability (%)
IVDMD (%)
Dry season
Edible biomass production
(DM, kg/ha)1
CP (%)
CP degradability (%)
IVDMD (%)
Losses due to trampling (%)
Rainy season
Dry season
Residual biomass from
rainy to dry
season (DM, kg/ha)2
Jaragua
Brachiaria
Arachis
Cratylia
10
10
10
15
3.2
4.5
Sugarcane
10
1.6
8
50
45
10
60
60
20
70
50
18
60
50
2
20
60
640
900
4000
3
50
30
4
60
35
20
70
50
18
60
50
2
20
60
25
20
800
25
20
1125
25
20
250
0
0
2000
0
0
16,000
1. Equivalent to 20% of biomass production during rainy season for grasses and 100% for cratylia.
2. Equivalent to 25% of biomass production during rainy season for grasses and 100% for cratylia and sugarcane.
Situation prevailing during study period

The average farm was assessed as to management practices and use of resources based on
the results of each individual country survey (Tables 1 to 3).
Level 1
This level consisted in the adoption of new germplasm to substitute purchased feeds
used to supplement the herd, especially during the dry season (Table 3). The same herd
size (the same number of milking cows) and milk production per cow were assumed for
this scenario. This adoption level requires the lowest level of investment and new forage
alternatives will be established in small areas. With this level of adoption no areas are
released for other alternative uses because its sole objective is to partially substitute purchased feed supplements.
119
Level 2
This level implied the adoption of new alternatives to not only satisfy substitute feeds
purchased to supplement the herd, but also to release areas currently allocated to livestock production for other alternative uses. As in Level 1, the same herd size and milk
production per cow were assumed; this level required a slightly higher investment than
Level 1 and could be considered the most rational option for producers already in Level
1.
Level 3
This level of adoption consisted in the establishment of new forage germplasm in the
entire farm area allocated to livestock production. This level of adoption assumed the
same milk production per cow but allowed the expansion in herd size. Therefore this
level required the highest level of investment because it not only required that larger
areas be established with improved forages, but also that additional animals be purchased to maximise land use. This level could be the option for producers already in
Level 2.

On-farm analysis of benefits
Tables 5, 6 and 7 present the results of simulation runs of the linear programming
model to analyse the situation at that time versus potential benefits of three levels of
adoption of new forage germplasm in dual-purpose farms in Costa Rica, Honduras and
Nicaragua. The situation of the dual-purpose farms surveyed was different in each of the
three countries. Nicaragua presented the lowest milk productivity per hectare because
practically all the farms surveyed were covered by naturalised Jaragua pastures (Table 7).
The opposite situation was observed in Honduras where milk productivity per hectare
was highest because the farmers surveyed had significant areas under improved pastures
(>60%), and important areas under cut-and-carry systems, for example sugarcane and
forage sorghum (Table 6). The situation observed in Costa Rica was intermediate, with
most of the pasture area covered by naturalised pastures and a low level of adoption of
improved forages (<10% of the area under livestock production). In all countries most
farms supplemented herds with purchased feeds, especially during the dry season. The
amount invested in supplement feeds/cow per year was highest in Honduras (US$ 96),
lowest in Nicaragua (US$ 32), and intermediate in Costa Rica (US$ 71).
The production cost/kg milk was lowest in Honduras (US$ 0.16), intermediate in
Nicaragua (US$ 0.18) and highest in Costa Rica (US$ 0.21). Differences in production
costs obeyed labour cost, which was more expensive in Costa Rica (US$ 9.30/personday)
120
and less expensive in Nicaragua (US$ 2.30/personday), but also the herd size and/or
stocking rate, which was 1.44 AU/ha in Honduras and 0.75 AU/ha in Costa Rica.
Table 5. Use of a simulation model to compare the situation of dual-purpose cattle farms in the dry
tropics of Costa Rica vs. the adoption of new forage germplasm to eliminate the purchase of feeds and to
improve production efficiency.
Parameter
Adult cows (no.)
Land use (hectare)
Native pasture
Sugarcane
Cratylia argentea
Brachiaria spp. alone
Brachiaria spp. + Arachis pintoi
Total area (hectare)
Additional investment for new
germplasm (US$)
Per farm
Per cow
Use of purchased feed (kg/cow
per year)
Chicken manure
Feed concentrates
Total cost (US$/cow per year)
Net annual income
Per hectare
Per cow
Per farm
Milk production (kg/ha)
Use of labour (no. of cows/person)
Area released
No. of hectares
Percent of livestock area
Milk production cost (US$/kg)
Prevailing
situation
Level 1a
Level 2b
Level 3c
23
23
23
48
40.5
0.3
0
7.2
0
48
38.5
0.7
1.6
7.2
0
48
0
1.0
1.7
26.5
0
29.2
0
4.9
3.6
0
39.5
48
na
na
0.75
872
38
0.75
5730
249
1.23
0
0
0
0
0
0
491
270
71
16
34
782
666
17.3
na
na
0.21
43
89
2047
666
15.9
0
0
0.18
98
124
2852
1095
18.5
18.8
39.2
0.16
15,960
333
1.56
0
0
0
119
119
5712
1390
17.6
na
na
0.16
a. Level 1 = Substitution of purchased feed.

b. Level 2 = Substitution of purchased feed + release of area.
c. Level 3 = Establishment of new forages on the farm.
na = not available.
Level 1
This level of adoption aimed to establish minimal areas of improved to replace the use of
purchased feeds for the herd, especially during the dry season. The forage option that
121
achieved this goal most economically was the establishment of Cratylia argentea as protein
source together with sugarcane as energy source.
Table 6. Use of a simulation model to compare the situation of dual-purpose systems in the dry tropics
of Honduras versus the adoption of new forage germplasm to eliminate the purchase of feeds and to
improve production efficiency.
Parameter
Adult cows (no.)
Land use (hectare)
Native pasture
Sugarcane
Cratylia argentea
Brachiaria spp. alone
Total area
Additional investment in new
germplasm (US$)
Per farm
Per cow
Stocking rate (AU/hectare)
Use of purchased feed (kg/cow per
year)
Molasses
Feed concentrates
Net annual income
Per hectare
Per cow
Per farm
Area released
No. of hectares
Prevailing
situation
Level 1a
Level 2b
Level 3c
42
42
42
46
13.7
3.2
0
27.1
0
44
10.4
3.4
3.0
27.1
0
44
0
3.7
2.9
33.3
0
39.9
0
4.1
3.2
36.7
0
44
na
na
1.44
850
20
1.44
1830
44
1.59
2580
56
1.58
0
0
0
0
0
0
0
0
0
88
418
96
160
167
7014
1386
15.3
249
261
10,962
1386
14.8
na
na
0.16
0
0
0.12
276
264
11,088
1530
15.2
4.1
9.3
0.11
277
265
12,190
1530
15.1
na
na
0.11

na = not available.
122
Table 7. Use of a simulation model to compare the situation of dual-purpose systems in the dry tropics of Nicaragua
versus the adoption of new forage germplasm to eliminate the purchase of feeds and to improve production efficiency.
Prevailing
Parameter
situation
Level 1a
Adult cows (no.)
44
44
Land use (hectare)
Native pastures
80
77.7
Sugarcane
0
0.6
Cratylia argentea
0
1.7
Brachiaria spp.
0
0
0
0
Total area
80
80
Additional investment in new
germplasm (US$)
Per farm
na
695
Per cow
na
16
0.99
0.99
Use of purchased feed (kg/cow per
year)
Chicken manure
410
0
Molasses
70
0
32
0
Net annual income (US$)
Per hectare
21
38
Per cow
38
69
Per farm
1672
3036
493
493
13.2
12.9
Area released
No. of hectares
na
0
na
0
0.18
0.16
na = not available.
Level 2b
Level 3c
44
86
0
4.4
2.5
37.3
0
44.2
0
9.2
5.2
0
65.6
80
7410
168
1.80
0
0
0
91
91
4044
892
14.7
35.8
44.8
0.14
22,070
256
1.95
0
0
0
93
87
7482
964
14.9
na
na
0.14
In the case of Costa Rica, the establishment of 0.7 ha of sugarcane with 1.6 ha of
Cratylia represented an investment of US$ 872 for a herd of 23 cows. This area produced enough nutrients and biomass to completely eliminate the need to purchase 491
kg of chicken manure and 270 kg of feed concentrate usually required per cow each year.
This investment was equivalent to US$ 38/cow, and generated savings in cash flow of
US$ 71/cow per year, allowing an increase in annual net income/cow of US$55 (from
US$ 34 to 89) and a reduction in production cost/kg milk of US$ 0.03 (from US$ 0.21
to 0.18).
The potential benefit of the new technology was even greater in Honduras than in
Costa Rica because the annual cost of purchased feeds in Honduras was the highest of
123
the three countries. Therefore, with a small investment of US$ 850/farm to establish 3.4
ha of sugarcane and 3 ha of Cratylia, the purchase of 88 kg of molasses and 418 kg of
feed concentrate/cow per year could be eliminated (Table 6). In addition, this level of
adoption reduced production cost/kg milk by US$ 0.04 (from US$ 0.16 to 0.12) and
increased the net income/cow per year by US$ 94 (from US$ 167 to 261).
The economic impact of this level of technology in Nicaragua was similar to that observed in Honduras. In this case, small areas (0.6 ha of sugarcane + 1.7 ha of Cratylia)
were needed to replace the use of purchased feeds that amounted to 410 kg of chicken
manure and 70 kg of molasses per cow/year, with a very modest investment of US$ 695/
farm (Table 7). The investment to achieve this level of adoption reduced production
cost/kg milk by US$ 0.02 (from US$ 0.18 to 0.16), which increased net income/cow per
year by US$ 31 (from US$ 38 to 69).
Level 2
This level of adoption aimed not only to substitute purchased feeds by establishing new
forage options, but also to release areas currently allocated to livestock production for
other alternative uses. To meet this objective, the simulation model incorporated not
only the forage alternatives used in Level 1, but also the adoption of improved grasses
such as Brachiaria spp.
In the case of Costa Rica, this level of adoption required, in addition to the investment of Level 1 and the 7.2 ha of improved grasses already existing, that another 19.3 ha
of Brachiaria spp. be established. This investment cost US$ 5730/farm (US$ 249/cow)
and allowed up to 18.8 ha of land, equivalent to 39.2% of the area allocated to livestock
production, to be released for alternative uses. As a result, the stocking rate was increased from 0.75 AU/ha to 1.23 AU/ha. The adoption of improved forages reduced
milk production costs even more than in Level 1 and increased net income (Table 5).
The additional investment to reach this level of adoption in Honduras was lower
than that required in Costa Rica: US$ 1830/farm (US$ 44/cow) because the farms
surveyed reported, on average, the existence of 27.1 ha of improved grasses. Thus, to
achieve the objectives of this level of adoption, an investment similar to that of Level 1
was required, and producers also had to invest in the establishment of an additional 6.2
ha of Brachiaria spp.
Because the stocking rate in Honduras was relatively high (approximately 1.44 AU/
ha) as a result of the adoption of improved grasses and cut-and-carry forages, the area
that could be released without affecting herd size and production/farm was low compared with Costa Rica and Nicaragua. The area that could have been released in
Honduras was about 4.1 ha, which represented 9.3% of the total farm area allocated to
livestock production (Table 6). This level of adoption led to additional reduction in
production cost/kg milk (US$ 0.11/kg) and an increase in net income/cow per year.
Level 2 required a higher level of investment in Nicaragua than in Costa Rica and
Nicaragua because the areas to be established with improved forages were larger in view
of the precarious situation of farms at that time. To achieve Level 2 of adoption, an
124
investment of US$ 7410/farm, equivalent to US$ 168/cow, was necessary to establish

4.4 ha of sugarcane, 2.2 ha of Cratylia, and 37.3 ha of Brachiaria spp. (Table 7). This
investment increased the stocking rate from 1 AU/ha to 1.80 AU/ha, and released
about 35.8 ha of land for other alternative uses (44.8% of area allocated to livestock
production). With this level of adoption, it was possible to reduce even more milk
production costs and increase net income/cow per year.
Level 3
This level attempts to show the potential of adopting new forage germplasm in the entire
farm area allocated to livestock production. This level of adoption required the largest
amount of capital investment because not only were improved forages established in
larger areas, but also more animals had to be purchased as a result of increased stocking
rate. This level of adoption can be regarded as the most intensive of the three levels
considered in this study, both in terms of productivity/hectare and in terms of capital
and labour use. The simulation model included the legume Arachis pintoi associated with
grasses of Brachiaria spp.
In Costa Rica this level of adoption required the additional establishment of 3.6 ha
of Cratylia and 4.9 ha of sugarcane, in relation to levels 1 and 2, and the establishment
of A. pintoi associated with Brachiaria in all pasture area (39.5 ha). The investment was
US$ 15,960/farm, equivalent to US$ 333/cow, which allowed a significant increase in
stocking rate (1.56 AU/ha) and the herd size (from 23 to 48 cows). This increase in herd
size generated other benefits, such as increased milk productivity/hectare during the
survey: from 666 kg milk/hectare at Level 1 to 1095 kg milk/hectare at Level 2 and 1390
kg milk/hectare at Level 3.
Although the annual net income/cow and per hectare varies little regarding Level 2
(increased from US$ 98 to 119/hectare but decreased from US$ 124 to 119/cow), the
net income/farm increased markedly because of the effect of stocking rate and expansion of herd size, practically duplicating the annual net income of producers from US$
2852/farm with Level 2 to US$ 5712 with Level 3.
As in Costa Rica, this level of adoption in Honduras required an additional increase
in the area planted to Cratylia (3.2 ha) and sugarcane (4.1 ha), and the establishment of
A. pintoi associated with Brachiaria (36.7 ha) in the total area under pastures. The investment to achieve this level of adoption totalled US$ 2580/farm and was significantly
lower than that required in Costa Rica because the farms surveyed in Honduras already
had more than 60% of the pasture area established with improved grasses, including important areas of cut-and-carry forages (3.2 ha or more per farm). This level of adoption
allowed farmers in Honduras to increase herd size by 9.5%, from 42 to 46 cows, thus
increasing net income (US$ 12,190/farm) as compared with the other levels of adoption
(Table 6).
The case of Nicaragua was similar to that of Costa Rica and Honduras because this
level requires an additional increase in the area planted to Cratylia (for example, from
2.5 ha to 5.2 ha) and to sugarcane (for example, from 4.4 ha to 9.2 ha), and the estab-
125
lishment of A. pintoi associated with Brachiaria spp. in the totality of the area under
pastures (65.6 ha). The investment to achieve this level of adoption in Nicaragua was the
highest of the three countries studied (US$ 22,070/farm) because the farms surveyed in
Nicaragua were completely covered with naturalised pastures. This investment in Nicaragua allowed producers to double the current stocking rate from 0.99 AU/ha to 1.95
AU/ha and increase herd size from 44 to 86 cows, thus increasing net income to US$
7482/farm per year, which is significantly higher than that obtained at Level 2 (US$
4044/farm per year). As a result, even though the investment to reach this level of
adoption is high, so is the potential benefit regarding the other two countries.
As in Costa Rica and Honduras, milk production cost and net income/cow and per
hectare in Nicaragua at Level 3 were similar to those of Level 2; however, net income/
farm increased by 85% because of the increase in herd size.
Regional benefit analysis

Table 8 presents the area under pasture; the number of livestock producers, and the livestock population located in the dry tropics of each region where the germplasm validated
by the Tropileche Consortium could be successfully replicated. Table 9 shows the potential benefits of new forage germplasm for each level of adoption, assuming these are effectively adopted in accordance with the data reported in Table 8.
Table 8. Characteristics of dual-purpose farms in the dry tropics of Costa Rica, Honduras, and
Nicaragua and adoption potential of new forage germplasm (Tropileche Consortium).
Characteristic
Costa Rica1
Area in pasture (ha)
528,254
Producers (no.)
18,768
Cattle population (no. of cows) 167,323
Honduras2
Nicaragua3
364,852
18,722
133,642
819,590
21,447
320,380
Sources: 1. = SEPSA (1990); 2. = Direccin General de Estadsticas y Censos (1998); 3. = MAG

(1999).
Level 1
This level of adoption had the largest impact on the situation of farmers because large
increases in net income/cow per year were obtained (from US$ 31 to 94) with small
investments per cow (between US$ 16 and 38). Cash flow increased with the substitution of purchased feed supplements with improved germplasm.
The investment required in each region ranged from US$ 2.7 million in Honduras to
US$ 6.4 million in Costa Rica. This investment was represented by the establishment of
new areas under improved forages, and estimated on the basis of the cost of each forage
alternative (Table 9).
The analysis of benefits, in terms of annual net income in cash for farmers in the
region, increased from US$ 9.9 million to 12.6 million, which surpassed the one-time
126
investment required every 10 years. The payback period necessary to return the investment with the marginal income was lower or equal to 6 months in all countries. In other
words, the initial investment could be paid back in the following 6 months, once the
improved forages had been established, before the onset of the dry season, when feed
supplements would be needed.
Table 9. Potential benefits of new forage germplasm at three levels of adoption in dual-purpose
cattle farms in the dry tropics of Costa Rica, Honduras and Nicaragua.
Level of adoption1
Country
Costa Rica
Investment/cow
Increase in income/cow per year2
Investment at regional level (US$ 106)
Increase in regional income/year (US$ 106)2
Years to pay investment
Honduras
Investment/cow
Nicaragua
Investment/cow
38
71
6.4
11.9
0.5
249
90
41.7
15.1
2.8
333
85
55.7
29.6
1.9
20
94
2.7
12.6
0.2
44
97
5.9
12.9
0.5
56
98
8.2
14.3
0.6
16
31
5.1
9.9
0.5
168
53
53.8
17.0
3.2
256
49
82.0
30.7
2.7
1. Level 1 implied the adoption of new forage alternatives with the minimum area needed to
eliminate the use of purchased feeds to maintain the same milk production and herd size.
Level 2 implied the adoption of new alternatives to satisfy Level 1 and to also release the largest
area under pasture for other alternative uses. Level 3 implied the adoption of improved
forages on the entire farm to achieve the highest production potential both on-farm and at the
regional level.
2. Marginal increase regarding the average situation of farms surveyed in each country.
Level 2
This level, as analysed before, should be the next phase once Level 1 has been achieved
because it demands a higher investment of capital, and is therefore not easy to adopt
initially.
The capital required for this level of adoption in different regions and countries
ranged between US$ 6 and 54 million, equivalent to an investment of US$ 44/cow in
Honduras and US$ 249/cow in Nicaragua. Furthermore, the payback period varied
127
from 6 months to 3.2 years because the levels of investment were highest, except in
Honduras where there was already a significant adoption of improved forages.
Level 3
This level was the most difficult because it demanded the highest investment. Achieving
an impact at the regional level required from US$ 8 million in Honduras to US$ 82
million in Nicaragua; however, the benefit, in terms of additional income for producers
in the dry tropics in each country, ranged between US$ 14.3 million/year in Honduras
and US$ 30.7 million/year in Nicaragua. Thus, based on this cash flow, the payback
period needed to recover the initial investment varied from 0.6 years in Honduras to 2.7
years in Nicaragua.
Seed requirements for technology at the regional

level
Table 10 presents seed production needs required for the establishment of each forage
alternative, according to the level of adoption at the regional level in each country (Table
8), assuming an annual rate of adoption of 10% of the available area for each alternative.
Level 1
About nine tonnes of C. argentea seed are required per country per year. Seed production/hectare of Cratylia is approximately 160 kg/ha (Argel and Lascano 1998) and
the demand for seed is about 8 kg/ha. Because availability of Cratylia seed is a limitation,
to satisfy the demand it will be necessary to establish 58 ha of seed multiplication plots
in Costa Rica, 48 ha in Honduras, and 62 ha in Nicaragua.
The demand of sugarcane propagation material is similar to that of Cratylia. It is
estimated that 1 ha of sugarcane produces 60 t of planting material for multiplication
and that approximately 14 t are needed to establish 1 ha. Therefore to satisfy planting
needs, 68 ha of seed multiplication plot must be established in Costa Rica, 15 ha in
Honduras, and 102 ha in Nicaragua.
Level 2
The amount of seed required to achieve this level of adoption at the regional level is
greater because the area to plant is larger, especially of Brachiaria spp. However, grass
seed is not a constraint because of the existence of local seed companies.
To meet the seed needs of C. argentea, this level of adoption required the establishment of 62 ha of additional seed multiplication plots in Costa Rica, 48 ha in Honduras and 92 ha in Nicaragua. To meet the seed requirements of sugarcane, 119 ha of
128
Table 10. Annual seed requirements of each forage species needed to achieve each level of adoption, estimating an
annual impact on 10% of the livestock population, in Costa Rica, Honduras and Nicaragua.
Costa Rica
Type of germplasm 1
and level of adoption
Level 1
Brachiaria spp.
Cratylia argentea
Arachis pintoib
Sugarcane
Level 2
Brachiaria spp.
Cratylia argentea
Arachis pintoi2
Sugarcane
Level 3
Brachiaria spp.
Cratylia argentea
Arachis pintoib
Sugarcane
Honduras
Area to
Seed
plant required
(ha)
(t/year)
Nicaragua
Area to
plant
(ha)
Seed
required
(t/year)
Area to
plant
(ha)
Seed
required
(t/year)
0
1164
0
291
0
9.3
0
4074
0
955
0
64
0
7.6
0
896
0
1240
0
437
0
9.9
0
6118
14,040
1237
0
509
56.2
9.9
0
7129
1970
950
0
160
7.9
7.6
0
2240
27,160
1820
0
3204
108.6
14.6
0
44,850
11,260
1260
13,770
1605
45.0
10.1
110.2
22,450
2790
930
0
261
11.2
7.4
0
3654
24,440
1940
24,440
3430
97.8
15.5
195.5
48,000
1. Level 1 implied the adoption of new forage alternatives with the minimum area needed to eliminate the use of
purchased feeds to maintain the same milk production and herd size. Level 2 implied the adoption of new
alternatives to satisfy Level 1 and to also release the greatest amount possible of area under pasture for other
alternative uses. Level 3 implied the adoption of improved forages in the entire farm area to reach the highest
production potential both on farm and at the regional level.
2. It is assumed that Arachis pintoi is established in association with an improved grass.
seed multiplication plots must be established in Costa Rica, 38 ha in Honduras and 748
ha in Nicaragua.
Level 3
This level of adoption required amounts of seed similar to Level 2, but it also required
high amounts of seed of A. pintoi to establish associations with Brachiaria. As in the case
of grass seed, the availability of Arachis seed did not represent a constraint because there
were sufficient quantities in the regional market.
To meet the requirements of Cratylia seed for this level of adoption, an additional 64
ha of seed multiplications plots must be established in Costa Rica, 47 ha in Honduras,
and 97 ha in Nicaragua, whereas for sugarcane 374 ha are needed in Costa Rica, 61 ha
in Honduras, and 800 ha in Nicaragua.
129
Strategy to promote the adoption of new

technologies
The requirements to reach a level of adoption similar to Level 1 are relatively easy to
achieve given the large benefits that can be obtained in terms of reduced production
costs and increased cash flow due to the substitution of purchased inputs with improved
germplasm. Use of family labour, frequently under-utilised or with a very low opportunity cost, especially in Nicaragua and Honduras, is more efficient.
The strategy to ensure successful adoption of Level 1 and subsequent advance toward
other levels of adoption consists in establishing strategic alliances with organised farmers
groups and/or extension agencies. These alliances allow the:
identification of producers interested in establishing seed multiplication plots on
their farms so that they become facilitators for the local expansion of new areas in
each region, and
creation of a follow-up mechanism to find markets for the seed produced, by identifying early adopters.
Conclusions
The potential benefits that result from the progressive adoption of new forage options
based on grasses and legumes validated by the Tropileche Consortium include the following:
Level 1 of adoption (new germplasm based on cratylia and sugarcane to substitute
feed supplements for herds) allows producers in all three countries to completely
eliminate the need to purchase feed supplements for the herd, especially during the
dry season. This effect has a large impact on the cash flow of producers because milk
production cost can be reduced by 14% in Costa Rica, by 25% in Honduras, and by
11% in Nicaragua.
Level 2 of adoption (areas of cratylia and sugarcane similar to Level 1, plus variable
areas of Brachiaria in each country) not only offers the benefits obtained with Level 1,
but also allows the producer to maintain the same milk production and herd size in a
smaller area, thus releasing areas to be allocated to other alternative uses. The released areas vary from 9% in the case of Honduras to 39% in Costa Rica and 45% in
Nicaragua. In addition, this level of adoption reduces milk production cost as compared with Level 1 by an additional 11% in Costa Rica, 8% in Honduras, and 12%
in Nicaragua.
Level 3 of adoption (areas of cratylia and sugarcane larger than in Levels 1 and 2, plus
pastures of Brachiaria +Arachis in association) presents the most intensive productivity and requires the highest level of capital investment. The highest milk productivity/hectare is achieved with this level of adoption1390 kg/ha in Costa Rica,
1530 kg/ha in Honduras, and 964 kg/ha in Nicaraguaand the highest stocking
rate1.56 AU/ha in Costa Rica, 1.58 AU/ha in Honduras and 1.95 AU/ha in
Nicaragua.
130
The potential impact that can be achieved by adopting these technologies at the re-
gional level is significant. The required amount of investment to achieve Level 1 in

Costa Rica was US$ 6.4 million; in Honduras, US$ 2.7 million; and in Nicaragua,
US$ 5.1 million. Because of the reduced production cost, it was possible to obtain an
additional annual net income in all regions studied: US$ 11.9 million in Costa Rica,
US$ 12.6 million in Honduras, and US$ 9.9 million in Nicaragua. Thus, the
potential benefit is quite large regarding the investment, and could be paid back in
less than one year (Table 9).
The investment required at the regional level to reach level 2 of adoption amounted
to US$ 41.7 million in Costa Rica, US$ 5.9 million in Honduras, and US$ 53.8
million in Nicaragua. The marginal income per year expected by the reduction in
production cost amounted to US$ 15.1 million in Costa Rica, US$ 12.9 million in
Honduras, and US$ 17.0 million in Nicaragua.
The resources required to obtain Level 3 of adoption were estimated at US$ 55.7
million in Costa Rica, US$ 8.2 million in Honduras, and US$ 82 million in Nicaragua, and the additional annual income that can be obtained by reduction in production cost and higher milk sales represented US$ 29.6 million in Costa Rica, US$
14.3 million in Honduras, and US$ 30.7 million in Nicaragua.
To promote the adoption of these technologies at the regional level, a strategy must
be designed to establish alliances with organised farmers groups and/or extension
agencies, and a mechanism to supply seed to producers, especially of C. argentea and
sugarcane. The availability of seed of Brachiaria and A. pintoi does not represent a
constraint.
This study suggests that large economic incentives exist to design a regional strategy
in each country to promote the adoption of technologies at Level 1 to improve the
income of small milk producers and increase the competitiveness of dual-purpose
production systems through the use of new improved forage germplasm based on
grasses and legumes.
References
Argel P. and Lascano C. 1998. Cratylia argentea: Una nueva leguminosa arbustiva para suelos
cidos en zonas subhmedas tropicales. Pasturas Tropicales 20(1):3743.
Direccin General de Estadsticas y Censos. 1998. Encuesta agrcola nacional 19961997. Tomo
III. Ganadera. Secretara de Industria y Comercio. Tegucigalpa, Honduras.
Holmann F., Kerridge P.C. and Lascano C.E. 1999. Progress Report. Improved legume-based
feeding systems for smallholder dual-purpose cattle production in tropical Latin America.
CIAT (Centro Internacional de Agricultura Tropical), Cali, Colombia.
MAG (Ministerio de Agricultura y Ganadera). 1999. Resultados de la encuesta de produccin de
1997. Convenio MAG-USDA. Managua, Nicaragua.
SEPSA (Secretara Ejecutiva de Planificacion Sectorial Agropecuaria). 1990. Encuesta Ganadera
Nacional 1988. San Jos, Costa Rica.
Umaa V. 1998. Comercio de productos lcteos en Centroamrica. Instituto Interamericano de
Cooperacin para la Agricultura (IICA). San Jos, Costa Rica.
131
Analysis of milk market of small-scale

artisan cheese factories in watersheds
of Honduras and Nicaragua involved
in livestock production
F. Holmann
CIATILRI, Colombia
Processing plants collect 20% of the milk produced in Honduras and Nicaragua that is
turned into pasteurised milk and other dairy products. The price paid for milk during
the study period was approximately US$ 0.30/kg in both countries, but this price was
obtained by less than 5% of dairy farms in both countries because these plants require
that milk be cooled to improve its hygienic quality and that farms be located in easily
accessible sites with good roads (Argel 1999a, 1999b).
The informal market collects the remaining 80% of the milk produced in both
countries, mainly in the form of cheese (Argel 1999a, 1999b). This market is mainly
constituted by small-scale artisan cheese factoriesmostly located in milk-producing
areasthat do not pasteurise milk. These factories transform milk into popular fresh
cheeses with shelf lives shorter than 10 days (de Franco et al. 1995). In Honduras alone,
there were about 600 artisan cheese factories at the time of the study (Argel 1999a). As a
result, the artisan cheese industry in both countries is the principal buyer of milk produced by small and medium producers who do not fulfil the requirements of milk processing plants of the known formal sector.
Seasonality in production
The price paid to producers by artisan cheese factories depends on the supply of milk,
which has a marked seasonality: it is abundant during the rainy season and scarce during
the dry season. This situation leads to serious difficulties in allocating surplus milk produced during the rainy season because both countries lack an efficient industrialisation
and marketing system. The seasonality in production causes milk prices to fluctuate,
with differences of up to 50% between seasons (Cajina 1994; de Franco et al. 1995).
In addition to this problem of seasonality, most milk collected by artisan cheese factories during the rainy season is of poor quality with bacterial counts close to 1 million/
cc. The pasteurisation process does not correct this problem for raw milks with bacterial
counts higher than 500 thousand bacteria/cc because pasteurisation only guarantees the
132
Analysis of milk market of artisan cheese factories in Honduras and Nicaragua
elimination of pathogenic bacteria, and many others remain alive thus affecting cheese
quality (de Franco et al. 1996).
Therefore it is not possible to make cheese of acceptable quality with this type of
milk, a situation determined by the sanitary conditions during milking and the handling
of milk until it reaches the cheese factories. Most milk producers in Honduras and
Nicaragua do not have milking sheds or clean, running water. Milking is carried out in
open corrals, by hand, using open buckets and surrounded by sludge. Milk cans are
washed with any type of common detergent and chlorine is not always used to disinfect
them (de Franco et al. 1996).
The milk produced under these conditions during the dry season is of higher quality
because there is no excess humidity in the corrals, the udders of cows are cleaner, and
the transportation time to the rural cheese factories is shortened because roads are in
better conditions (Garca 1996).
As a result of these changes in supply and quality, it is more attractive to produce
higher volumes of milk during the dry season because it benefits the producer, who receives a higher price, and the artisan cheese factories, that receive milk of better quality.
Objective
This study was done to analyse the milk market of small-scale artisan cheese factories
located in the livestock production areas of Honduras and Nicaragua where the
Tropileche Consortium operates. The study aimed to answer the following questions:
Is there a market for additional milk production? If so, how much milk can be
absorbed by artisan cheese factories each season?
Is there a market for milk of higher hygienic quality and, if so, would the farmgate
price be higher? What price would the cheese factories is willing to pay?

The data for this study was obtained through surveys carried out during March 2000 to
10 small-scale artisan cheese factories in Honduras and 13 in Nicaragua. These cheese
factories were located in the area of influence of the Tropileche Consortium: Olancho,
Catacamas, and Juticalpa in Honduras, and in Esquipulas and Muy-Muy in Nicaragua.
Secondary information from both countries was used to complement survey data.

Collection of milk and prices paid
The average milk collected was about 6 times more in Honduras than in Nicaragua in
both rainy (6200 kg/day vs. 1100 kg/day) and dry season (3600 kg/day vs. 520 kg/day)
133
(Tables 1 and 2). Milk collected during the rainy season was 73% more than that collected during the dry season in Honduras and 111% more in Nicaragua. This factor
disrupts any market and has significant impact on the price of milk that artisan cheese
factories are willing to pay. Therefore, the milk price during the rainy season compared
with that paid during the dry season was 27% lower in Honduras and 38% lower in
Nicaragua. In addition, the milk price paid by cheese factories for raw material of a similar quality was 33% higher in Honduras than in Nicaragua during the rainy season
(US$ 0.24/kg vs. US$ 0.18/kg) and 14% higher during the dry season (US$ 0.33/kg vs.
US$ 0.29/kg).
Table 1. Characteristics of milk and cheese production and marketing of 10 small-scale rural artisan cheese factories
during the rainy and dry seasons in the watersheds of Olancho, Catacamas, and Juticalpa (Honduras) dedicated to
livestock production.
Characteristic
Gathering of milk (litre/day)
Purchase price (US$/litre)
Types of cheese produced and sale price (US$/kg)
Dry cheese
Fresh cheese
Cream cheese
String cheese (quesillo)
Cream
Amount of milk required to produce each type of
cheese (litre or kg)
Dry cheese
Fresh cheese
Cream cheese
String cheese (quesillo)
Cream
Percentage of cheese processors willing to purchase
more milk
Percentage of processors who considered that the
milk was of good quality
Yes
No
Higher price consumers were willing to pay for
better quality milk
US$/litre
Percent of processors
Dry season
3600
0.33
Rainy season
6230
0.24
3.60
2.70
2.40
2.09
2.40
2.40
1.80
1.94
1.30
1.80
9.48
6.00
5.73
5.62
14.70
55.50
10.14
6.76
6.39
6.45
16.90
0
10
90
10
90
0.361
9.40
0.266
11.20
Types of cheeses and yields

Four types of cheese are produced in both countries by artisan factories. They also sell
cream as a by-product because most cheeses are made with skimmed milk. The sale of
cream represents a net profit in both countries and each type of cheese has a different
134
price. Therefore, the marketing strategy was similar, being segmented to four different
types of cheese (tastes) associated with the purchasing power of consumers.
Table 2. Characteristics of milk and cheese production and marketing of 13 small-scale rural
artisan cheese factories, during the rainy and dry seasons, in the watersheds of Esquipulas and
Muy-Muy (Nicaragua) dedicated to livestock production.
Characteristics
Collection of milk (litre/day)
Purchase price (US$/litre)
Types of cheese produced and sale price (US$/kg)
Pure cheese
Media Sangre cheese
Cottage cheese
Cream cheese
Cream
Amount of milk required to produce each type of
cheese (litre/kg)
Pure cheese
Media Sangre cheese
Cottage cheese
Cream cheese
Cream
Percentage of cheese processors willing to purchase
more milk
Percentage of cheese processors who considered
that the milk was of good quality
Yes
No
Higher price consumers were willing to pay for
better quality milk
US$/litre
Percent of processors
Dry season
Rainy season
523
0.29
1103
0.18
2.49
2.13
1.78
1.78
1.60
1.78
1.60
1.24
1.24
0.89
7.71
6.61
5.51
5.51
11.03
75.70
8.82
7.71
6.61
6.61
13.23
0
80
20
70
30
0.29
0
0.211
17.00
Although all types of cheese produced are fresh and have short shelf lifes, the difference between them is the amount of moisture they contain. To produce a cheese with
low moisture content, a larger amount of fluid milk is required and, as a result, its price
is higher (for example, dry cheese in Honduras and pure cheese in Nicaragua). This situation is demonstrated by the amount of fluid milk necessary to produce 1 kg of cheese,
which, in turn, is directly related to sale price. The higher the moisture contents of the
cheese, the shorter its shelf life. Therefore, low-cost cheeses are also those with the shortest shelf life because they contain more whey.
The amount of milk necessary to produce each type of cheese varies according to the
season of the year. During the rainy season, milk production/cow is higher but its solid
contents is lower and, as a result, more milk is required to produce the same amount of
cheese. For example, in Honduras the artisan factories required 7% more milk during
135
the rainy season to produce 1 kg of dry cheese and up to 15% more milk to produce 1
kg of string cheese or quesillo. In Nicaragua, artisan factories required 14% more milk
during the rainy season to produce pure cheese and up to 19% more milk to produce 1
kg of cream cheese. This reduction in yield efficiency during the rainy season forces
artisan factories to pay a lower milk price to producers.
Potential market and milk quality

The potential growth of the cheese market was large and unsatisfied. The artisan cheese
factories surveyed in Honduras could purchase up to 55% more milk during the dry
season, but there was no milk on offer. In Nicaragua the potential was even greater because artisan factories were willing to purchase up to 76% more milk than that collected
during the time of the study.
This situation is not the same during the rainy season, when the potential for growth
is null in both countries because of the excess offer of milk in the market. As a result,
the artisan factories in both countries would only be willing to purchase more milk during the dry season, which has significant implications for the type of technologies that
should be promoted among producers in both regions.
In Honduras, 90% of artisan factories surveyed considered that the milk collected
during the rainy season had low hygienic quality, but this figure decreased to only 10%
during the dry season. Therefore there was a direct relationship between low hygienic
quality of milk and rainy season, which agrees with de Franco et al. (1996). In Nicaragua,
30% of artisan factories considered that the milk produced during the rainy season was
of low hygienic quality, which is still significant even though lower than in Honduras.
Artisan cheese factories in Honduras and Nicaragua that considered that the milk
collected was of low hygienic quality would be willing to pay a higher price if the quality
of the milk improved. In Honduras this price would be 9.4% higher during the dry
season and 11.2% higher during the rainy season. In Nicaragua the artisan cheese factories would be willing to pay 17% more, but only during the rainy season.
Needs of small-scale artisan cheese factories

Table 3 presents the needs and expectations of artisan cheese factories in both countries.
The main need was the improvement of factory infrastructure by acquiring more equipment (for example, creameries, cold rooms, devices to determine hygienic quality). The
future prospects of the cheese industry in both countries are promising.
The improvement of milk quality was ranked as second priority in Honduras, while
the search for new markets was ranked second in Nicaragua (Table 3). Nicaragua began
to export fresh cheeses to El Salvador and Honduras in 1990 for a gross value of US$
128 thousand. Since then, exports have been on the rise, reaching 8400 t of cheese in
1998 for a gross value of US$ 14 million (MAGFOR 2000). All factors considered
indicate that this trend will continue in coming years.
136
Table 3. Needs of rural artisan cheese factories in Honduras and Nicaragua (% of cheese factories).
Need
More equipment to expand infrastructure1
Improved hygienic quality of milk
Technical assistance to produce new types of cheese

Improved quality of cheeses
Expansion of existing markets and search for new markets
Honduras
(%)
Nicaragua
(%)
80
62
60
15
50
15
40
23
10
46
1. Acquisition of creameries, cold rooms, stainless steel tubs, and/or burners.

2. Technical assistance to improve hygienic conditions at milking and reduce transportation time of
milk from farm to cheese factory.
3. Includes improving hygienic conditions of the artisan cheese factory and packaging of product.
The third need identified in Honduras was the lack of technical assistance to produce
new types of cheeses and thus diversify the market, and not so much the search for new
markets for the same types of cheeses. In Nicaragua, the need was to improve the quality
of cheeses produced by artisan factories.
Technological implications
The results of this survey have two types of technological implications for the Tropileche
Consortium: animal feeding and genetic improvement of animals.
Regarding animal feeding, these results suggest that an aggressive programme to promote the shrub legume Cratylia argentea and sugarcane as supplements of cattle herds
during the dry season would have a much higher impact than the promotion of grasses
or legumes during the rainy season. This technological change would reduce the need to
purchase feed concentrates to supplement the herd, thus improving the cash flow of
producers and increasing milk fat content.
The protein contained in milk is the main input to produce cheese. In other words,
milk with higher protein content yields more cheese. It is estimated that for each 0.1%
increase in protein content, cheese yield is increased by 4% (Sozzi 1999).
Therefore, the other technological implication of improving yield efficiency of cheese
making is by using milk with higher protein content, and this is mainly achieved by
genetic improvement of animals at the farm level by crossbreeding with cows that produce milk with higher protein content. The type of livestock commonly found in
Honduras and Nicaragua is Zebu (Brahman). The protein content of Bos indicus breeds
such as Brahman, Guzerath and Nelore is 3.10%, with little variability among them. In
Bos taurus breeds variability is greater, ranging from 3.15% for Holstein cows, to 3.50%
for Brown Swiss cows, to 3.80% for Jersey cows (Ruiz 1999). As a result, a genetic improvement strategy that incorporates Jersey or Brown Swiss genes into the local Zebu
gene pool would allow producers to offer artisan cheese factories milk with higher
protein content, provided that a transparent payment system that values milk protein
exists.
137
Conclusions
Study results and the implications of this analysis led to the following conclusions:
The main buyer of the milk produced by small and medium producers in Honduras
and Nicaragua is the small-scale artisan cheese industry.
The production of milk in watersheds in the dry tropics where the Tropileche
Consortium operates is seasonal.
Production during the rainy season is practically twice that of the dry season, causing
over-supply during the rainy season and scarcity of milk during the dry season.
The milk shortage during the dry season in both countries leads to an unsatisfied
market potential.
Of artisan factories in Honduras, 55.5% would be willing to buy more milk during
the dry season, and of those in Nicaragua, 75.7%, but this supply is not available
because of the lack of adoption of technologies for animal feeding based on low-cost
improved forages.
The above situation suggests that an aggressive programme to promote the use of
shrub legume C. argentea with sugarcane to supplement the milking herd during the
dry season would have greater impact than the promotion of grasses and legumes for
the rainy season. This technological change would reduce the need to purchase feed
concentrates to supplement the herd, thus improving the cash flow of producers and
increasing milk fat content.
Artisan cheese factories in both countries, but especially in Honduras, require higher
quality milk, particularly during the rainy season. In Honduras these factories would
be willing to pay 9.4% more during the dry season and 11.2% more during the rainy
season. In Nicaragua the artisan factories would be willing to pay 17% more, but only
during the rainy season.
The main need of artisan cheese factories was the improvement of infrastructure by
purchasing more equipment. Future prospects for the industry seem promising in
both countries. The second most important need in Honduras was to produce and
purchase milk of higher hygienic quality, while in Nicaragua the second most important need was the search for new markets. The third most important need in
Honduras was the lack of technical assistance to produce new types of cheese to
diversify the market, while in Nicaragua it was the improvement of the quality of
cheese produced.
The amount of milk necessary to produce each type of cheese varied depending on
the season of the year. During the rainy season, milk production/cow was higher but
contained less total solids and, as a result, more milk was required to yield the same
amount of cheese as during the dry season.
References
Argel P. 1999a. Caracterizacin de sitios con potencial de adopcin de germoplasma forrajero
mejorado por pequeos y medianos productores de doble propsito en Honduras. Informe de
Consultora. CIAT (Centro Internacional de Agricultura Tropical), Cali, Colombia.
138
Argel P. 1999b. Caracterizacin de sitios con potencial de adopcin de germoplasma forrajero

mejorado por pequeos y medianos productores de doble propsito en Nicaragua. Informe de
Consultora. CIAT (Centro Internacional de Agricultura Tropical), Cali, Colombia.
Cajina A. 1994. Precios y mercado de la leche en Nicaragua. Unidad de Programa Agrcola
MAG/BID/PNUD. Managua, Nicaragua.
de Franco M., Sacasa N., Raudes R., Sevilla R. and Ziga L. 1995. La dinmica del mercado de
productos lcteos en Nicaragua. Anlisis Total. Consultores en Gerencia y Economa.
Managua, Nicaragua.
de Franco M., Sevilla R., Downing H. and Sandino A. 1996. Caracterizacin y propuesta de
estrategia para el sector semi-industrial de productos lcteos en Nicaragua. Anlisis Total.
Consultores en Gerencia y Economa. Managua, Nicaragua.
Garca L. 1996. Resultados de la encuesta a la agroindustria de quesos artesanales en cinco
departamentos de Nicaragua. Proyecto de Desarrollo Lechero, Ministerio de Agricultura y
Ganadera (MAG), Managua, Nicaragua.
MAGFOR (Ministerio Agropecuario y Forestal). 2000. Agricultura y Desarrollo. Servicio de
Informacin de Precios y Mercados no. 56. Ministerio Agropecuario y Forestal, Managua,
Nicaragua.
Ruiz F. 1999. Mejoramiento gentico de los componentes de la leche. In: Memorias del II
Seminario Internacional sobre Calidad de Leche: Competitividad y Protena. Cooperativa
Lechera COLANTA, Medelln, Colombia.
Sozzi T. 1999. Importancia de las protenas en la leche y sus derivados. In: Memorias del II
Seminario Internacional sobre Calidad de Leche: Competitividad y Protena. Cooperativa
Lechera COLANTA, Medelln, Colombia.
139
Research results
Component 3
Usefulness of new forage systems
Activity 3.2 Dissemination of research results

Highlights
The dissemination of results is an integral part of the research and development
process and is essential in the adoption of new technologies.
Tropileche Newsletter
F. Holmann
CIATILRI, Colombia, and national institutions in Peru, Costa Rica, Nicaragua and Honduras
The Tropileche Consortium has published 10 issues of the Tropileche Newsletter.

Publication dates are March and October. This newsletter aims to inform our partners
about the activities of the Consortium, on-going research trials, research results at the
different benchmark sites, and any other news that could prove useful for our readers.
The printed newsletter is distributed to those interested and can also be consulted on
Internet at the Tropileche website: http:/www.ciat.cgiar.org/tropileche/start.htm.
143
Database on dual-purpose cattle

production systems
CIATILRI, Colombia
In October 1996, the Tropileche Consortium developed a database that contains the
results of research conducted on dual-purpose cattle systems in tropical Latin America
since 1960. Topics covered are nutrition and feeding, forages (grasses and legumes),
genetic improvement of animals and reproduction, animal health, economics and
extension, transfer and adoption of technology.
The database currently contains about 2260 references, all with basic descriptors;
about 70% also include an abstract. This database was developed in Micro CD/ISIS and
follows the guidelines of FAOs AGRISCARIS information system.
The database can be accessed on Internet at the Tropileche website
(http:/www.ciat.cgiar.org/tropileche/start.htm). Over the last two years, an average of
139 users consulted the database each month.
144
Tropileche in the Internet

CIATILRI, Colombia
The Tropileche Consortium has developed its own web site, which contains the
newsletters produced and the database on research results generated in tropical Latin
America. Users can access the website at http:/www.ciat.cgiar.org/tropileche/start.htm,
by clicking on the icon Soils and Systems or through the Information and
Documentation Unit.
The website also offers a list of researchers in dual-purpose cattle production systems
in Latin America and the Caribbean, with contact addresses. All those interested can
access the database and read about the Consortiums activities. Users may also request
information, and communicate and interact with other colleagues.
145
Videotape release
F. Holmann,1 C. Lascano,1 P. Argel1 and R. Goyenaga2
1. CIAT, Colombia
2. MAG, Costa Rica
As part of Tropileches strategy to diffuse information about technology adoption by

farmers, the Consortium, with the collaboration of the Communications Department of
the Ministry of Agriculture of Costa Rica, has produced an 11-minute videotape on the
evolution of a dual-purpose farm, owned by Antonio Lpez, in the dry tropics of the
Central Pacific region of Costa Rica.
Antonio is a smalholder farmer and one of the Consortiums collaborators who have
adopted many of the technologies promoted by the Consortium through Costa Ricas
Ministry of Agriculture and Livestock Production (MAG). Antonio is currently producing more milk on less area; he has doubled the family income, and has released areas
from livestock production, which are now protecting the farms water sources.
This videotape will be used to show other producers in Costa Rica and elsewhere in
Latin America how one smallholder farmer succeeded in intensifying production with
new forage technologies, based on improved grasses and legumes, with the technical
assistance of MAG and seed donated by the Tropileche Consortium.
146
Publication of brochures
As part of a strategy to disseminate information on the new technologies being adopted
by producers, the Tropileche Consortium, in collaboration with the Communications
Department of Costa Ricas Ministry of Agriculture and Livestock Production, published
two brochures for agricultural extension agents. These brochures aimed to explain, in
simple terms, the knowledge acquired about the establishment and management of the
shrub legume Cratylia argentea and the grass Brachiaria brizantha CIAT 26110 cv. Toledo,
and their impact on livestock productivity. Summaries of these publications follow.
147
Cratylia argentea: A shrub legume

for livestock production in the tropics
P. Argel,1 J. Gonzalez2 and M. Lobo3
1. CIAT, Colombia
2. ECAG
3. MAG, Costa Rica
Cratylia argentea (synonyms = C. floribunda, Dioclea floribunda) is a shrub legume that is

evaluated and selected by the MAG, ECAG, UCR and CIAT for its adaptability to
prolonged dry seasons and acid soils in Costa Rica.
Description
Cratylia argentea is a shrub that grows naturally in Brazil. Its distribution ranges, in a
northsouth direction, from the State of Par to the States of Mato Grosso and Gois
and, in an eastwest direction, from the State of Cear to Peru. The shrub is between
1.5 m and 3.0 m high when it grows in the open, but becomes a climbing plant when
associated with larger plants.
Adaptation
Cratylia argentea grows naturally at altitudes between 0 and 930 metres above sea level
(masl), although most inhabit areas between 300 and 800 masl. In Costa Rica, this
legume adapts well to a broad range of climates and soils, particularly to subhumid
climates and to acid Ultisols, which are poor soils with high aluminium content. The
plant also responds to fertile soils, and the highest yields have been reported from fertile
sites in the humid tropics. Nevertheless, the shrub does not tolerate poorly drained sites
or heavy soils that are frequently saturated in rainy seasons. The shrub has an outstanding capacity to regrow and shows high foliar retention during dry seasons. A major
reason is the development of vigorous roots that can be as long as 2 m. Thus, between
30 and 40% of the total DM yield can be produced during that critical period of the
yearthe dry season.
Pests and diseases

So far, neither pests nor diseases of economic importance have been reported to attack
C. argentea. At some sites, isolated cases of plant death have been observed during estab-
148
Cratylia argentea: A shrub legume for livestock production in the tropics
lishment. The causal agents were jobotos or dung beetles (Melolonthidae spp.) in the soil,
crickets and leaf-cutting ants. For adult plants, isolated cases of death have also been
reported from sites with heavy soils and frequent rains. Roots of infected plants have
been found to carry fungi of the genera Pythium and Fusarium, although these were not
proven to be the direct cause of mortality. In other similar cases of plant death, nematodes were found near the stem crown. Pod blight, caused by the fungal genera Phoma
and Cladosporium, has been observed when flowering and fruiting occur under rainy
conditions with high relative humidity.
Planting
Cratylia argentea propagates easily through seed, whereas vegetative propagation through
stakes has, so far, not been successful. Because seeds are soft, they do not need scarifying,
but they need to be sown close to the soils surface, that is, no more than 2 cm deep.
Deep sowing causes seed rot, retards seedling emergence, and produces plants with less
developed root systems. Sowing may be direct into the field with minimum tillage or
after conventional land preparation with plough and harrow. Seeds may also be sown in
bags for later transplanting to the field. Depending on soil fertility, fertilisation is,
ideally, with phosphorus (100 to 150 kg/ha of triple superphosphate).
Cratylia argentea seeds respond to inoculation with the strains of the cowpea type of
Rhizobium, which are very common in tropical soils. Recent experiments show that the
legume responds well to effective nodule formation by Rhizobium strains, CIAT 3561 and
3564, particularly in acid soils with high aluminium content.
Cratylia argentea grows slowly, at least during the first two months after sowing, even
though its seedlings are more vigorous than those of other shrubby legumes.
Seed production and quality

Flowering in C. argentea is abundant but poorly synchronised and starts at the end of the
rainy season in the lowland tropics. Plants can flower the first year they are established,
but seed yields are low. Flowering lasts three months and pollinating insects are commonly seen around the flowers. First fruits are mature about one and a half months after
pollination, and fruiting continues through the next 2 or 3 months. Harvesting seed is
therefore a continuous process, carried out manually, once a week, throughout most of
the dry season (February and March).
Seed yield depends on genotype, plant age, cutting management and prevailing environmental conditions during flowering and fruiting. In Atenas, Costa Rica, 3-year-old
plants, cut at 30 cm height, and fertilised with phosphorus at the beginning of the rains,
yield, on the average, 50 to 70 g of pure seed per plant (CIAT 1999). However, the date
of the uniformity cut affects first flowering and, as a result, the potential seed yield.
Plants cut toward the end of the rainy season or within the dry season tend to flower
little, forming few fruits.
149
Cratylia argentea seed does not have marked physical (due to hardness of seed coat) or
physiological dormancy, and can lose its viability relatively quickly if it is stored under
the temperature and humidity conditions prevailing in the lowland tropics.
Nutritive value
The usable forage of 3-month-old C. argentea (leaves + thin and tender stems) has a CP
content of 23%, but this figure varies from 19% to 26% according to the plants age.
Similarly, the in vitro dry matter digestibility of C. argentea ranges between 40 and 55%,
which is higher than that of other shrub legumes adapted to acid soils, such as
Codariocalyx giroides (30%) and Flemingia macrophylla (20%). The high percentage of CP
and the low content of condensed tannins found in C. argentea make this legume an
excellent source of nitrogen for the ruminant (Wilson and Lascano 1997).
Use and management

Cratylia argentea has shown that it is an excellent protein supplement in diets for dairy
cows grazing poor quality grasses during the dry season. The highest response has been
from cows with medium and high dairy potential and fed the legume as fresh and
chopped or as ensiled, together with high-energy sources such as sugarcane. For example,
for Jersey cows, as much as 66% of commercial concentrate can be replaced by C. argentea,
either fresh or ensiled, without significant loss in milk production. The ensiled diet
tends toward higher fat than does the fresh-legume diet (Romero and Gonzalez 2000).
Similar results have been reported for dual-purpose cows, which were offered a daily diet
of 12 kg of chopped sugarcane, 6 kg of chopped or ensiled C. argentea, and 0.6 kg of rice
polishing (Table 1) (Lobo and Acua 2000). Of the three supplements, fresh C. argentea
was the most economical, because of its lower costs and the cows significantly higher
milk production.
Table 1. Average milk production of dual-purpose cows given supplements of either chicken manure or of the legume
Cratylia argentea, offered either fresh or ensiled, during the dry season on a farm in Esparza, Costa Rica.
Diet1
Ensiled Cratylia
Fresh Cratylia
Chicken manure
Milk
(kg/cow per day)
Fat
(%)
Cost of supplement/
cow per day (US$)
Cost:benefit
ratio
5.1
5.5
5.3
3.6
3.4
3.0
0.164
0.109
0.217
1.6
2.4
1.1
1. Chicken manure, sugarcane and rice polishing were each offered at 3 kg/cow per day. Fresh C. Argentea
contained 20% CP and 1.8 Mcal of ME, while the ensiled legume had 16.5% CP and 1.9 Mcal ME. The pH
for ensiled Cratylia was 4.5 and digestibility was 36%.
Cratylia argentea can also be used in direct grazing (browsing), with the legume either
established as a protein bank or planted in bands within the paddocks. Experiment results showed that grazing cows having access to a bank of C. argentea consumed more
mature foliage and less immature forage. A major advantage of this legume, therefore, is
150
Cratylia argentea: A shrub legume for livestock production in the tropics
that, in direct grazing systems, its use by ruminants can be deferred to the dry season. No
other management practices need be followed.
In evaluations where cuts were made every 8 weeks during the rainy season and every
12 weeks during the dry season, C. argentea responded very well. Adult plants that were
more than 1 year old had the highest yield and quality of forage when cut at 90 cm and
after the respective 60 days regrowth (Lobo et al. unpublished data). Total yield can range
between 2 and 4 t DM/cut, although this depends on sowing density and the plants age.
A sowing density of 1.0 0.5 m gave the highest yield/hectare when planting C. argentea
as protein bank. The legumes good response to cutting shows the high potential that it
has for cut-and-carry systems.
Limitations
Cratylia argentea does not adapt well to climates found above 1200 maslat least, the
most advanced ecotypes being evaluated (CIAT 18516 and CIAT 18668) do not. This
legume establishes slowly, but more quickly than do other shrub legumes. The forage
yields are also low during the first year of establishment.
References
Lobo Di Palma M. and Acua R. 2000. Efecto de la suplementacin con Cratylia argentea cv.
Veraniega fresca y ensilada en vacas de doble propsito en el trpico subhmedo de Costa
rica. In: Holmann F. and Lascano C. (eds), Final Report Tropileche Project. CIAT (Centro
Internacional de Agricultura Tropical) and ILRI (International Livestock Research Institute).
(in press).
Romero R.F. and Gonzlez J. 2000. Efecto de la alimentacin durante la poca seca con Cratylia
argentea fresca y ensilada sobre la produccin de leche y sus componentes. In: Holmann F.
and Lascano C. (eds), Final Report Tropileche Project. CIAT (Centro Internacional de
Agricultura Tropical) and ILRI (International Livestock Research Institute). (in press).
Wilson Q.T. and Lascano C.E. 1997. Cratylia argentea como suplemento de un heno de gramnea
de baja calidad utilizado por bovinos. Pasturas Tropicales 19(3):28.
151
Brachiaria brizantha CIAT 26110:

cv. Toledo
P. Argel,1 J. Gonzalez2 and M. Lobo3
1. CIAT, Colombia
2. ECAG
3. MAG, Costa Rica
In October 2000, the Costa Rican Ministry of Agriculture and Livestock Production
(MAG) released the new cultivar Toledo (Brachiaria brizantha CIAT 26110) for farmers
use. This material was officially released during a field day in which 250 farmers, professionals, and seed producers of the region participated. Research that led to the development of this new cultivar began in 1988 with the introduction of experimental seed
from the International Centre for Tropical Agriculture (CIAT) in Colombia. The research was co-ordinated by the Tropileche Consortium in Costa Rica, with the participation of MAG, the Escuela Centroamericana de Ganadera (ECAG), the Tropical
Centre for Agricultural Research and Training (CATIE), the University of Costa Rica
and CIAT.
Origin and morphology

The accession Brachiaria brizantha CIAT 26110 was collected in Burundi, on 15 May
1985 by G Keller-Grein of CIAT in collaboration with technicians from the Institut des
Sciences Agronomiques du Burundi (ISABU). The collection site was at 2 53 S, 26 20
W, at an altitude of 1510 metres above sea level (masl), and an average annual rainfall of
1710 mm. The site was located 36 km between Bubanza and Bukinanyama in the State
of Cibitoke. In October of the same year, this accession was registered as CIAT 26110 in
the germplasm bank held at CIAT. The accession was introduced into Costa Rica in
1988 for evaluation with other Brachiaria species at Los Diamantes experiment station in
Gupiles, according to the co-operative agreement between MAG, CATIE, ECAG and
CIATs former Tropical Forages Program.
Cultivar Toledo was derived directly from the B. brizantha accession CIAT 26110. It
is a perennial grass that develops in form of tillers, and can grow as high as 1.6 m. It produces vigorous stalks that root at the nodes if these should come into close contact with
the soil through, for example, trampling by animals or mechanical compaction, thus
favouring soil cover and lateral displacement of the grass. Leaves are lanceolate, with
little pubescence, and can be 60 cm long and 2.5 cm wide. The inflorescence is a panicle,
40 to 50 cm long. It usually carries four racemes, ranging from 8 to 12 cm in length and
152
Brachiaria brizantha CIAT 26110: cv. Toledo
each carrying a single row of spike lets. One stalk can produce several inflorescences,
each growing from a different node, although the largest is always at the terminal.
Adaptation and dry matter yields

Cultivar Toledo adapts to a broad range of climates and soils, and grows well, not only
in subhumid tropical sites with a 56 month dry season but also in humid tropical sites
with an average annual rainfall of 4300 mm. This characteristic was observed in the agronomic trials carried out at 11 sites in Colombia by the Colombian Brachiaria Evaluation
Network. Although the grass grows well in less fertile and acid soils, it performs better at
sites where soils are of medium to good fertility. It tolerates sandy soils and poorly-drained
sites, although growth is reduced if the soil is waterlogged for more than 30 days. It also
tolerates the dry season well, maintaining a higher proportion of green leaves than other
B. brizantha cultivars such as Diamantes 1 (Marand) and La Libertad. The higher greenery is probably a result of the high leaf tissue content of non-structural carbohydrates
(197 mg/kg of DM) and few minerals (8% of ash) in cultivar Toledo. This cultivar also
grows well under shade and, in Costa Rica, at mid- altitudes (1500 masl) where temperatures average 18 C.
Biomass production of cultivar Toledo was high in Inceptisols in Costa Rica and
Panama, located at sites with contrasting climate, both during the rainy and dry season.
Average production of cultivar Toledo across 11 contrasting sites in Colombia, with cuts
every 8 weeks, was about 3.88 t DM/ha for the dry season and 5.1 t DM/ha for the rainy
season. These yields were higher than those of other Brachiaria accessions evaluated at
the same sites and under similar management conditions.
Planting
Cultivar Toledo is easily established by means of its seed, which are of good quality and
give rise to vigorous seedlings. It can also be propagated vegetatively but, for best results,
splits should be used.
Seeds can be broadcast or sown in furrows that are spaced at 0.6 m. The land may be
conventionally prepared with plough and harrow, or the cover vegetation is controlled
with non-selective herbicides according to zero tillage practices. The amount of seed to
use depends on its quality (purity and germination rate) and on the sowing method. For
example, sowing in ploughed and harrowed furrows requires smaller quantities of seed,
compared with sowing by broadcast on land receiving zero or minimal tillage. The final
quantity of seed used can range between 3 and 4 kg/ha for seed classified at 60% (e.g.
80% purity and 75% germination rate).
The high vigour of seedlings of cultivar Toledo and their initial aggressive growth
allows this cultivar to compete adequately with weeds during establishment. Consequently, the collective experiences of farmers already growing this cultivar in Costa Rica
is that the grass can be first grazed 3 to 4 months after sowing.
153
Nutritive value and animal production

Evaluations carried out at ECAG in Atenas showed that CP contents of leaves are 13.5,
10.1 and 8.7% for the regrowth ages of 25, 35 and 45 days, respectively. The in vitro dry
matter digestibility values for the same ages were 67.8, 64.2 and 60.3%, respectively.
These results suggest that cultivar Toledo has similar or slightly better forage qualities
than other B. brizantha cultivars.
Observations made on farms in Costa Rica and information provided by cattle producers indicate that cultivar Toledo can support a stocking rate between 2.5 and 3.0
AU/ha during the rainy season, with a frequency of grazing every 2130 days. One dualpurpose farm reported that daily milk production tended to be similar to that observed
for pastures of other Brachiaria species in association with the forage peanut (Arachis
pintoi). Initial reports from Colombia indicate that the daily milk production of a
Holstein cow is significantly less on cultivar Toledo (8.0 litre/cow) than on B. decumbens
cv. Pasto Peludo or B. brizantha cv. Diamantes 1 (8.8 and 8.9 litre/cow, respectively).
Use and management

So far, cultivar Toledo has been used only for grazing cattle, although horses have been
seen to consume tender leaves. Although little is known about the most adequate occupation and rest periods, considering its rapid recovery, an adequate rest period could
be between 21 and 28 days. In addition, this cultivars high forage production allows
stocking rates higher than 2.5 AU/ha, at least during the rainy season.
Because of its growth habit in form of tillers, this cultivar associates well with stoloniferous legumes such as Arachis pintoi, resulting in better soil cover and improved forage
quality. These observations are still being evaluated in Costa Rica. Although caltivar
Toledo is adequate for grazing, because of its size and vigorous growth, it could be successfully used in cut-and-carry forage systems.
154
Publications
Book chapters
Argel P.J. 1999. Man forrajero. Una leguminosa de uso mltiple para el sector agropecuario de
Costa Rica. Montecillos (Costa Rica) 15(102):1213.
Argel P.J. and Paton C.J. 1999. Overcoming legume hardseedness. In: Loch D.S. and Ferguson
J.E. (eds), Forage seed production, Volume 2: Tropical and subtropical species. CAB (Commonwealth Agricultural Bureau) International, Wallingford, UK. pp. 247265.
Holmann F. 1998. Evaluacin econmica de sistemas de produccin de leche en el trpico. In:
Vaccaro L. and Perez A. (eds), El desarrollo de la produccin de leche en Amrica Latina Tropical.
Archivos Latinoamericanos de Produccin Animal (ALPA) Supl 1:1931.
Holmann F. and Estrada R.D. 1997. Alternativas agropecuarias en la regin Pacfico Central de
Costa Rica: Un modelo de simulacin aplicable a sistemas de produccin animal de doble
propsito. In: Lascano C. and Holmann F. (eds), Concepto y metodologas de investigacin en
fincas con sistemas de produccin de doble propsito. CIAT (Centro Internacional de Agricultura
Tropical), Cali, Colombia.
Holmann F. and Lascano C. 1998. Una nueva estrategia para mejorar los sistemas de produccin
de doble propsito en los trpicos. El Consorcio Tropileche. In: Gonzalez-Stagnaro G.,
Madrid-Bury N. and Soto Belloso E. (eds), Mejora de la ganadera mestiza de doble propsito.
Maracaibo, Venezuela.
White D., Holmann F., Fujisaka S., Reategui K. and Lascano C. 2000. Will intensifying pasture
management in Latin America protect forestsor is it the other way around? In: Angelsen A.
and Kaimowitz D. (eds), Agricultural Technologies and Tropical Deforestation. CAB (Commonwealth Agricultural Bureau) International, Wallingford, UK.
Books
Lascano C. and Holmann F. 1997. Conceptos y metodologas de investigacin en fincas con sistemas de
produccin animal de doble propsito. CIAT (Centro Internacional de Agricultura Tropical), Cali,
Colombia. 285 pp.
Technical bulletins
Argel P.J. and Villareal M. 1998. Nuevo man forrajero perenne (Arachis pintoi Krapovickas and
Gregory). Cultivar Porvenir (CIAT 18744): leguminosa herbcea para alimentacin animal, el
mejoramiento y conservacin del suelo y el embellecimiento del paisaje. MAG (Ministerio de
Agricultura y Ganadera) and CIAT (Centro Internacional de Agricultura Tropical). Technical
Bulletin. 32 pp.
155
Argel P.J., Gonzalez J. and Lobo M. 2000. Cratylia argentea: una leguminosa arbustiva para la
ganadera del trpico. Ministerio de Agricultura y Ganadera, Costa Rica.
Argel P.J., Gonzalez J. and Lobo M. 2000. Brachiaria brizantha CIAT 26110: Cultivar Toledo.
Ministerio de Agricultura y Ganadera, Costa Rica.
Articles in scientific journals

Argel P.J. and Lascano C.E. 1998. Cratylia argentea (Desvaux) O. Kuntze: Una nueva leguminosa
arbustiva para suelos cidos en zonas hmedas tropicales. Pasturas Tropicales 20(1):3743.
Duarte O., Pulido J., Silva J. and Holmann F. 1999. Anlisis de la situacin actual y alternativas
tecnolgicas de los sistemas de produccin agropecuaria en el Valle del Cesar, Colombia.
Revista Corpoica (in press).
Holmann F. 1999. Ex ante evaluation of forage technologies in Peru, Costa Rica, and Nicaragua.
Journal of Livestock Research for Rural Development 11(3).
Holmann F. 2000. Beneficios potenciales de nuevo germoplasma forrajero en fincas de doble
propsito del trpico seco en Costa Rica, Honduras y Nicaragua. Journal of Livestock Research
for Rural Development (submitted).
Holmann F. 2001. Anlisis del Mercado de leche de queseras artesanales en cuencas ganaderas de
Honduras y Nicaragua. Journal of Livestock Research for Rural Development 13(1).
Holmann F., Peck D. and Lascano C. 2000. El dao econmico del salivazo de los pastos en
Colombia: Una primera aproximacin del impacto sobre la productividad animal en
Brachiaria decumbens. Journal of Economic Entomology (submitted).
Ibrahim M.A., Holmann F., Hernandez M. and Camero A. 2000. La contribucin de los bancos
de protena de Erythrina con desechos de bananas para el mejoramiento de los sistemas de
produccin en el trpico hmedo. Agroforestry Systems 49:245.
Pezo D.A., Holmann F. and Arze J. 1999. Evaluacin bioeconmica de un sistema de produccin
de leche basado en el uso intensivo de gramneas fertilizadas en el trpico hmedo de Costa
Rica. Revista Agronoma Costarricense 23(1):97109.
Rivas L. and Holmann F. 2000. Early adoption of Arachis pintoi in the humid tropics: The case of
dual-purpose systems in Caqueta, Colombia. Journal of Livestock Research for Rural Development
12(3).
Articles in popular magazines

Holmann F. and Lascano C. 1998. Utilidad de los sistemas de produccin en base a forrajes
mejorados. Revista Pecuaria de Nicaragua 2(20):17.
Holmann F. and Lascano C. 1998. Optimizacin de la produccin de forrajes. Revista Pecuaria de
Nicaragua 2(18):22.
156
Publications
Posters
Argel P.J., Lobo M., Gonzalez J., Romero F., Holmann F., Lascano C.E. and Kerridge P.C. The
leguminous shrub, Cratylia argentea: A dry season feeding alternative for the humid tropics.
Presented at the workshop Working with farmers: The key to adoption of forage-based
technologies, 1215 October, Philippines.
International seminars
Argel P.J. 1999. Tecnologas forrajeras para el desarrollo de una ganadera ms productivo en el
trpico bajo en Centroamrica. Contribucin del CIAT. In: Pomareda C. (ed), Proceedings of
the workshop on Intensificacin de la ganadera en Centroamrica: beneficios econmicos y ambientales.
FAO/CATIE, May 1999.
Argel P.J. and Perez G. 1998. Adaptation of new species of Leucaena in Costa Rica, Central
AmericaPreliminary results. In: Shelton H.M., Gutteridge R.C., Mullen B.F. and Bray R.A.
(eds), LeucaenaAdaptation, quality and farming systems. ACIAR Proceedings 86:313323.
ACIAR (Australian Centre for International Agricultural Research) Canberra, Australia. pp.
Argel P.J., Lascano C.E., and Ramirez L. 1998. Leucaena in Latin American farming systems:
Challenges for development. In: Shelton H.M., Gutteridge R.C., Mullen B.F. and Bray R.A.
(eds), LeucaenaAdaptation, quality and farming systems. ACIAR Proceedings 86:313323.
ACIAR (Australian Centre for International Agricultural Research) Canberra, Australia.
Estrada R.D., Holmann F. and Posada R. 2000. Farmer and industry funding of agricultural
research in Colombia. Paper presented at the symposium on Financing of Agricultural
Research. XXIV Conference of the International Association of Agricultural Economists
1318 August 2000. Berlin, Germany.
Holmann F. 1997. Competitividad de los distintos modelos de produccin de leche en Amrica
Latina tropical. Paper presented at the VI Congreso Panamericano de la Leche. FEPALE
(Federacin Panamericana de la Leche). April 1997. Buenos Aires, Argentina.
Holmann F. 1998. Estudio de opciones para el desarrollo e intensificacin de los sistemas de
produccin de leche en Amrica Latina tropical. Paper presented at the workshop on
Desarrollo de produccin de leche en Amrica Tropical, 1821 November 1998. Maracay,
Venezuela.
Holmann F. 2000. El uso de modelos de simulacin como herramienta para la toma de
decisiones en la promocin de nuevas alternativas forrajeras: El caso de Costa Rica y Per.
Paper presented at the XVI Meeting of the Latin American Association of Animal Production
(ALPA), 2831 March 2000. Montevideo, Uruguay.
Holmann F. and Lascano C. 1997. The contribution of Tropileche to the livestock research
agenda in Latin America. Annual Programme Meeting. ILRI (International Livestock
Research Institute). September 1997. Addis Ababa, Ethiopia.
Peters M., Horne P., Schmidt A., Holmann F., Kerridge P., Tarawali S.A., Schultze-Kraft R.,
Lascano C.E., Argel P., Str W., Fujisaka S., Mller-Smann K. and Wortmann C. 1999. The
role of forages in reducing poverty and degradation of natural resources in tropical
production systems. Paper presented at the Poverty Workshop, 1014 September 1999. San
Jos, Costa Rica.
157
White D., Holmann F., Fujisaka S., Reategui K. and Lascano C. 1999. La intensificacin a base
de pasturas mejoradas afecta el rea boscosa en Amrica Latina tropical? Invirtiendo la
pregunta. Paper presented at the international workshop on Tecnologas que Intensifican la
Produccin Agrcola y la Deforestacin, 1113 March 1999. CATIE, Turrialba, Costa Rica.
Workshop proceedings
Holmann F. 1998. Avances en las actividades del Consorcio Tropileche. ECAG (Escuela
Centroamericana de Ganadera), Atenas, Costa Rica, 2227 February 1998.
Holmann F., Lascano C. and Ramirez A. 1999. Workshop on advances of research on the
Tropileche Consortium. Moyobamba, Peru, 2830 June 1999.
Video
La finca de Antonio Lpez (11 minutes). CIAT and MAG, Costa Rica. August 1999.
Tropileche Newsletter
Published in October 1996; March and October 1997; March and October 1998; March and
October 1999; March and October 2000; and March 2001.
158
List of project proposals developed

Title
Evaluation of options to develop the dairy
sector of tropical Latin America
Integrating the conservation of biodiversity and
smallholder livestock systems in sub-humid
tropical landscapes
Improving dual-purpose livestock production
systems in Nicaragua
Improving the productivity of smallholder dairy
farms in Central America and the Caribbean
Improving dual-purpose livestock production
systems based on improved legumes in the
coastal areas of Ecuador
Developing new forage technologies for
smallholder dairies in the Alto Mayo Region of
the Peruvian Amazon
Amount requested
(US$ 103)
Duration
(years)
480
Donor
Fontagro
540
CRUSA & SLP
1940
Norway
1470
IDB
189
PROMSA/World
Bank
50
Peru
159
Tropileche collaborators
Name
Colombia
Federico Holmann
Carlos Lascano
Pedro Argel
Samuel Fujisaka
Anderson Medina
Patricia Avila
Alberto Ramrez
Jaime Velasquez
Costa Rica
Carlos Hidalgo
Marco Lobo
Vidal Acua
Francisco Romero
Jess Gonzlez
Carlos Jimenez
Muhammad Ibrahim
Peru
Keneth Reategui
Geiner Romero
Daisy Lara
Jorge Vela
Alfredo Riesco
Miguel Ara
Nicaragua
Tito Farias
Honduras
Conrado Burgos
160
Position
Institution
Co-ordinator, Tropileche Consortium

Animal nutritionist
Agronomist
Anthropologist
Administrative assistant
Research assistant
Editor
Agronomist
CIAT/ILRI
CIAT
CIAT
CIAT
CIAT
CIAT
CIAT
CORPOICA
National Co-ordinator
Agronomist
Agronomist
Animal nutritionist
Research assistant
Agronomist
Agronomist
MAG
MAG
MAG
ECAG
ECAG
UCR
CATIE
Agronomist
Research assistant
Agronomist
Agronomist
Economist
Agronomist
DEPAAM
CIAT
FUNDAAM
IIAP
CODESU (SDC)
IVITA
Agronomist
IDR
Agronomist
DICTA

Arachis ssp3 PDF

Diunggah oleh

Informasi Dokumen

Deskripsi Asli:

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Arachis ssp3 PDF

Diunggah oleh

Hak Cipta:

Format Tersedia

Feeding systems with forage legumes

to intensify dairy production in Latin

Feeding systems with forage

CGIAR System-wide Livestock Programme

International Center for Tropical Agriculture

International Livestock Research Institute

I. Lascano, Carlos. II. Tropileche Consortium. III. International Livestock Research

2004 ILRI (International Livestock Research Institute)

Research results: Component 2Activity 2.1

Feeding systems with forage legumes to intensify dairy production in LAC

Early adoption of Arachis pintoi in the humid tropics:

List of abbreviations and acronyms

acid detergent fibre

Feeding systems with forage legumes to intensify dairy production in LAC

Achievements attained for each component and activity of

1.1 Controlled feeding and

The use of forage legumes as protein supplements during

1.2 Use of forage resources

The response of cows in terms of milk production to

1.3 Functional relationships

Concentration of milk urea nitrogen (MUN) as an

Developing new 2.1 On-farm evaluation of

Forage systems for strategic feed supplementation during

2.2 Evaluating new feed

Fattening pre-weaned calves on legumes during the dry

Simulation models are useful for reducing research costs,

Feeding systems with forage legumes to intensify dairy production in LAC

3.1 Diagnosis, economic

Ex ante analyses of new improved technologies to

3.2 Disseminating research

The dissemination of results is integral to research and

Feeding systems with forage legumes to intensify dairy production in LAC

Optimising forage use

Activity 1.1 Controlled feeding and grazing trials

the milk yield of cows showing high genetic potential.

Effects of Brachiaria decumbens alone

Brachiaria decumbens alone, B. decumbens associated with A. pintoi and B. decumbens

Materials and methods

Feeding systems with forage legumes to intensify dairy production in LAC

docks to measure botanical composition, forage availability and quality of forage on

Brachiaria decumbens + Arachis pintoi

B. decumbens + Hyparrhenia rufa

Significance (P < 0.05)

* Grazing cycle every 3035 days.

Effects of B. decumbens alone and associated with A. pintoi on milk production

Brachiaria decumbens + Arachis pintoi

B. decumbens + Hyparrhenia rufa

1. 1 AU = 450 kg animal liveweight.

Feeding systems with forage legumes to intensify dairy production in LAC

Effects of B. decumbens alone and associated with A. pintoi on milk production

Milk production and composition

Brachiaria decumbens + Arachis

B. decumbens + Hyparrhenia rufa

Significance (P < 0.05)

Feeding systems with forage legumes to intensify dairy production in LAC

Effects of B. decumbens alone and associated with A. pintoi on milk production

Brachiaria decumbens + Arachis pintoi

B. decumbens + Hyparrhenia rufa

* Grazing cycles of 3035 days.

Rice (Oriza sativa) yield as an indicator

Stylosanthes guianensis (stylo) is a legume adapted to the humid rainforest conditions of

Materials and methods